+

Search Tips   |   Advanced Search


Monitor - WebSphere MQ v7.5

  1. Monitoring and performance
  2. Event monitoring
    1. Instrumentation events
    2. Controlling events
    3. Event queues
    4. Enabling queue depth events
    5. Configuration events
    6. Command events
    7. Logger events
  3. Message monitoring
  4. Activity report reference
  5. Trace-route message reference
  6. Accounting and statistics messages
  7. Application activity trace
  8. Real-time monitoring


Monitoring and performance

A number of monitoring techniques are available in WebSphere MQ to obtain statistics and other specific information about how your queue manager network is running. Use the monitoring information and guidance in this section to help improve the performance of your queue manager network.

Depending on the size and complexity of your queue manager network, you can obtain a range of information from monitoring your queue manager network. The following list provides examples of reasons for monitoring your queue manager network:


Event monitoring

Event monitoring is the process of detecting occurrences of instrumentation events in a queue manager network. An instrumentation event is a logical combination of events that is detected by a queue manager or channel instance. Such an event causes the queue manager or channel instance to put a special message, called an event message, on an event queue.

WebSphere MQ instrumentation events provide information about errors, warnings, and other significant occurrences in a queue manager. Use these events to monitor the operation of the queue managers in your queue manager network to achieve the following goals:


Instrumentation events

An instrumentation event is a logical combination of conditions that a queue manager or channel instance detects and puts a special message, called an event message, on an event queue.

WebSphere MQ instrumentation events provide information about errors, warnings, and other significant occurrences in a queue manager. You can use these events to monitor the operation of queue managers (with other methods such as Tivoli NetView for z/OS).

Figure 1 illustrates the concept of instrumentation events.

Figure 1. Understanding instrumentation events


Event monitoring applications

Applications that use events to monitor queue managers must include the following provisions:

  1. Set up channels between the queue managers in your network.
  2. Implement the required data conversions. The normal rules of data conversion apply. For example, if you are monitoring events on a UNIX system queue manager from a z/OS queue manager, ensure that you convert EBCDIC to ASCII.


Event notification through event queues

When an event occurs, the queue manager puts an event message on the appropriate event queue, if defined. The event message contains information about the event that you can retrieve by writing a suitable MQI application program that performs the following steps:

The related information describes the format of event messages.


Conditions that cause events

The following list gives examples of conditions that can cause instrumentation events:


Event types

Use this page to view the types of instrumentation event that a queue manager or channel instance can report

WebSphere MQ instrumentation events have the following types:

For each queue manager, each category of event has its own event queue. All events in that category result in an event message being put onto the same queue.

This event queue: Contains messages from:
SYSTEM.ADMIN.QMGR.EVENT Queue manager events
SYSTEM.ADMIN.CHANNEL.EVENT Channel events
SYSTEM.ADMIN.PERFM.EVENT Performance events
SYSTEM.ADMIN.CONFIG.EVENT Configuration events
SYSTEM.ADMIN.COMMAND.EVENT Command events
SYSTEM.ADMIN.LOGGER.EVENT Logger events
SYSTEM.ADMIN.PUBSUB.EVENT Gets events related to Publish/Subscribe

By incorporating instrumentation events into your own system management application, you can monitor the activities across many queue managers, across many different nodes, and for multiple WebSphere MQ applications. In particular, you can monitor all the nodes in your system from a single node (for those nodes that support WebSphere MQ events) as shown in Figure 1 .

Instrumentation events can be reported through a user-written reporting mechanism to an administration application that can present the events to an operator.

Figure 1. Monitoring queue managers across different platforms, on a single node

Instrumentation events also enable applications acting as agents for other administration networks, for example Tivoli NetView for z/OS, to monitor reports and create the appropriate alerts.


Queue manager events

Queue manager events are related to the use of resources within queue managers. For example, a queue manager event is generated if an application tries to put a message on a queue that does not exist.

The following examples are conditions that can cause a queue manager event:

WebSphere MQ puts messages for queue manager events on the SYSTEM.ADMIN.QMGR.EVENT queue, and supports the following queue manager event types:

Authority (on Windows, HP OpenVMS, and UNIX systems only)

Authority events report an authorization, such as an application trying to open a queue for which it does not have the required authority, or a command being issued from a user ID that does not have the required authority. The authority event message can contain the following event data:

  • Not Authorized (type 1)
  • Not Authorized (type 2)
  • Not Authorized (type 3)
  • Not Authorized (type 4)
  • Not Authorized (type 5)
  • Not Authorized (type 6)

All authority events are valid on HP OpenVMS, Windows, and UNIX systems only.

Inhibit

Inhibit events indicate that an MQPUT or MQGET operation has been attempted against a queue where the queue is inhibited for puts or gets, or against a topic where the topic is inhibited for publishes. The inhibit event message can contain the following event data:

  • Get Inhibited
  • Put Inhibited

Local

Local events indicate that an application (or the queue manager) has not been able to access a local queue or other local object. For example, an application might try to access an object that has not been defined. The local event message can contain the following event data:

  • Alias Base Queue Type Error
  • Unknown Alias Base Queue
  • Unknown Object Name

Remote

Remote events indicate that an application or the queue manager cannot access a remote queue on another queue manager. For example, the transmission queue to be used might not be correctly defined. The remote event message can contain the following event data:

  • Default Transmission Queue Type Error
  • Default Transmission Queue Usage Error
  • Queue Type Error
  • Remote Queue Name Error
  • Transmission Queue Type Error
  • Transmission Queue Usage Error
  • Unknown Default Transmission Queue
  • Unknown Remote Queue Manager
  • Unknown Transmission Queue

Start and stop

Start and stop events indicate that a queue manager has been started or has been requested to stop or quiesce.

z/OS supports only start events.

Stop events are not recorded unless the default message-persistence of the SYSTEM.ADMIN.QMGR.EVENT queue is defined as persistent. The start and stop event message can contain the following event data:

  • Queue Manager Active
  • Queue Manager Not Active

For each event type in this list, you can set a queue manager attribute to enable or disable the event type.


Concept Feedback


Last updated: Monday, September 23, 2013



Channel and bridge events

Channels report these events as a result of conditions detected during their operation. For example, when a channel instance is stopped.

Channel events are generated in the following circumstances:

Note: Client connections do not cause Channel Started or Channel Stopped events.

When a command is used to start a channel, an event is generated. Another event is generated when the channel instance starts. However, starting a channel by a listener, the runmqchl command, or a queue manager trigger message does not generate an event. In these cases, an event is generated only when the channel instance starts.

A successful start or stop channel command generates at least two events. These events are generated for both queue managers connected by the channel (providing they support events).

If a channel event is put on an event queue, an error condition causes the queue manager to create an event.

The event messages for channel and bridge events are put on the SYSTEM.ADMIN.CHANNEL.EVENT queue.

The channel event messages can contain the following event data:


SSL events

The only Secure Sockets Layer (SSL or TLS) event is the Channel SSL Error event. This event is reported when a channel using SSL or TLS fails to establish an SSL connection.

The SSL event messages can contain the following event data:


Concept Feedback


Last updated: Monday, September 23, 2013



Performance events

Performance events are notifications that a resource has reached a threshold condition. For example, a queue depth limit has been reached.

Performance events relate to conditions that can affect the performance of applications that use a specified queue. They are not generated for the event queues themselves.

The event type is returned in the command identifier field in the message data.

If a queue manager tries to put a queue manager event or performance event message on an event queue and an error that would typically create an event is detected, another event is not created and no action is taken.

MQGET and MQPUT calls within a unit of work can generate performance events regardless of whether the unit of work is committed or backed out.

The event messages for performance events are put on the SYSTEM.ADMIN.PERFM.EVENT queue.

There are two types of performance event:

Queue depth events

Queue depth events relate to the number of messages on a queue; that is, how full or empty the queue is. These events are supported for shared queues. The queue depth event messages can contain the following event data:

  • Queue Depth High
  • Queue Depth Low
  • Queue Full

Queue service interval events

Queue service interval events relate to whether messages are processed within a user-specified time interval. These events are not supported for shared queues.


Concept Feedback


Last updated: Monday, September 23, 2013



Configuration events

Configuration events are generated when a configuration event is requested explicitly, or automatically when an object is created, modified, or deleted.

A configuration event message contains information about the attributes of an object. For example, a configuration event message is generated if a namelist object is created, and contains information about the attributes of the namelist object.

The event messages for configuration events are put on the SYSTEM.ADMIN.CONFIG.EVENT queue.

There are four types of configuration event:

Create object events

Create object events are generated when an object is created. The event message contains the following event data: Create object .

Change object events

Change object events are generated when an object is changed. The event message contains the following event data: Change object .

Delete object events

Delete object events are generated when an object is deleted. The event message contains the following event data: Delete object .

Refresh object events

Refresh object events are generated by an explicit request to refresh. The event message contains the following event data: Refresh object .


Concept Feedback


Last updated: Monday, September 23, 2013



Command events

Command events are reported when an MQSC or PCF command runs successfully.

A command event message contains information about the origin, context, and content of a command. For example, a command event message is generated with such information if the MQSC command, ALTER QLOCAL, runs successfully.

The event messages for command events are put on the SYSTEM.ADMIN.COMMAND.EVENT queue.

Command events contain the following event data: Command .


Concept Feedback


Last updated: Monday, September 23, 2013



Logger events

Logger events are reported when a queue manager that uses linear logging starts writing log records to a new log extent.

A logger event message contains information specifying the log extents required by the queue manager to restart the queue manager, or for media recovery.

The event messages for logger events are put on the SYSTEM.ADMIN.LOGGER.EVENT queue.

The logger event message contains the following event data: Logger .


Concept Feedback


Last updated: Monday, September 23, 2013



Event message data summary

Use this summary to obtain information about the event data that each type of event message can contain.

Event type See these topics
Authority events Not Authorized (type 1)
Not Authorized (type 2)
Not Authorized (type 3)
Not Authorized (type 4)
Not Authorized (type 5)
Not Authorized (type 6)
Channel events Channel Activated
Channel Auto-definition Error
Channel Auto-definition OK
Channel Blocked
Channel Conversion Error
Channel Not Activated
Channel Started
Channel Stopped
Channel Stopped By User
Command events Command
Configuration events Create object
Change object
Delete object
Refresh object
IMS™ Bridge events Bridge Started
Bridge Stopped
Inhibit events Get Inhibited
Put Inhibited
Local events Alias Base Queue Type Error
Unknown Alias Base Queue
Unknown Object Name
Logger events Logger
Performance events Queue Depth High
Queue Depth Low
Queue Full
Queue Service Interval High
Queue Service Interval OK
Remote events Default Transmission Queue Type Error
Default Transmission Queue Usage Error
Queue Type Error
Remote Queue Name Error
Transmission Queue Type Error
Transmission Queue Usage Error
Unknown Default Transmission Queue
Unknown Remote Queue Manager
Unknown Transmission Queue
SSL events Channel SSL Error
Start and stop events Queue Manager Active
Queue Manager Not Active


Reference Feedback


Last updated: Monday, September 23, 2013



Controlling events

You enable and disable events by specifying the appropriate values for queue manager, queue attributes, or both, depending on the type of event.

You must enable each instrumentation event that you want to be generated. For example, the conditions causing a Queue Full event are:

Enable and disable events by using any of the following techniques:

Note: You can set attributes related to events for both queues and queue managers only by command. The MQI call MQSET does not support attributes related to events.


Controlling queue manager events

You control queue manager events by using queue manager attributes. To enable queue manager events, set the appropriate queue manager attribute to ENABLED. To disable queue manager events, set the appropriate queue manager attribute to DISABLED.

To enable or disable queue manager events, use the MQSC command ALTER QMGR, specifying the appropriate queue manager attribute. Table 1 summarizes how to enable queue manager events. To disable a queue manager event, set the appropriate parameter to DISABLED.

Enabling queue manager events using MQSC commands

Event ALTER QMGR parameter

Authority
Inhibit
Local
Remote
Start and Stop

AUTHOREV (ENABLED)
INHIBTEV (ENABLED)
LOCALEV (ENABLED)
REMOTEEV (ENABLED)
STRSTPEV (ENABLED)


Reference Feedback


Last updated: Monday, September 23, 2013



Controlling channel and bridge events

You control channel events by using queue manager attributes. To enable channel events, set the appropriate queue manager attribute to ENABLED. To disable channel events, set the appropriate queue manager attribute to DISABLED.

To enable or disable channels events use the MQSC command ALTER QMGR, specifying the appropriate queue manager attribute. Table 1 summarizes how you enable channel and bridge events. To disable a queue manager event, set the appropriate parameter to DISABLED.

Enabling channel and bridge events using MQSC commands

Event ALTER QMGR parameter

Channel
Related to channel errors only
IMS Bridge
SSL
Channel auto-definition

CHLEV (ENABLED)
CHLEV (EXCEPTION)
BRIDGEEV (ENABLED)
SSLEV (ENABLED)
CHADEV(ENABLED)


Reference Feedback


Last updated: Monday, September 23, 2013



Controlling performance events

You control performance events using the PERFMEV queue manager attribute. To enable performance events, set PERFMEV to ENABLED. To disable performance events, set the PERFMEV queue manager attribute to DISABLED.

To set the PERFMEV queue manager attribute to ENABLED, use the following MQSC command:

 ALTER QMGR PERFMEV (ENABLED)
To enable specific performance events, set the appropriate queue attribute. Also, specify the conditions that cause the event.

Queue depth events

By default, all queue depth events are disabled. To configure a queue for any of the queue depth events:

  1. Enable performance events on the queue manager.
  2. Enable the event on the required queue.
  3. Set the limits, if required, to the appropriate levels, expressed as a percentage of the maximum queue depth.

Queue service interval events

To configure a queue for queue service interval events you must:

  1. Enable performance events on the queue manager.
  2. Set the control attribute for a Queue Service Interval High or OK event on the queue as required.
  3. Specify the service interval time by setting the QSVCINT attribute for the queue to the appropriate length of time.

Note: When enabled, a queue service interval event can be generated at any appropriate time, not necessarily waiting until an MQI call for the queue is issued. However, if an MQI call is used on a queue to put or remove a message, any applicable performance event is generated at that time. The event is not generated when the elapsed time becomes equal to the service interval time.


Concept Feedback


Last updated: Monday, September 23, 2013



Controlling configuration, command, and logger events

You control configuration, command, and logger events by using the queue manager attributes CONFIGEV, CMDEV, and LOGGEREV. To enable these events, set the appropriate queue manager attribute to ENABLED. To disable these events, set the appropriate queue manager attribute to DISABLED.

Configuration events

To enable configuration events, set CONFIGEV to ENABLED. To disable configuration events, set CONFIGEV to DISABLED. For example, you can enable configuration events by using the following MQSC command:

 ALTER QMGR CONFIGEV (ENABLED)

Command events

To enable command events, set CMDEV to ENABLED. To enable command events for commands except DISPLAY MQSC commands and Inquire PCF commands, set the CMDEV to NODISPLAY. To disable command events, set CMDEV to DISABLED. For example, you can enable command events by using the following MQSC command:

 ALTER QMGR CMDEV (ENABLED)

Logger events

To enable logger events, set LOGGEREV to ENABLED. To disable logger events, set LOGGEREV to DISABLED. For example, you can enable logger events by using the following MQSC command:

 ALTER QMGR LOGGEREV(ENABLED)


Concept Feedback


Last updated: Monday, September 23, 2013



Event queues

When an event occurs, the queue manager puts an event message on the defined event queue. The event message contains information about the event.

You can define event queues either as local queues, alias queues, or as local definitions of remote queues. If you define all your event queues as local definitions of the same remote queue on one queue manager, you can centralize your monitoring activities.

You must not define event queues as transmission queues, because event messages have formats that are incompatible with the message format that is required for transmission queues.

Shared event queues are local queues defined with the QSGDISP(SHARED) value.

For more information about defining shared queues on z/OS, see Application programming with shared queues .


When an event queue is unavailable

If an event occurs when the event queue is not available, the event message is lost. For example, if you do not define an event queue for a category of event, all event messages for that category are lost. The event messages are not, for example, saved on the dead-letter (undelivered-message) queue.

However, you can define the event queue as a remote queue. Then, if there is a problem on the remote system putting messages to the resolved queue, the event message arrives on the dead-letter queue of the remote system.

An event queue might be unavailable for many different reasons including:

The absence of an event queue does not prevent the event from occurring. For example, after a performance event, the queue manager changes the queue attributes and resets the queue statistics. This change happens whether the event message is put on the performance event queue or not. The same is true in the case of configuration and command events.


Use triggered event queues

You can set up the event queues with triggers so that when an event is generated, the event message being put onto the event queue starts a user-written monitoring application. This application can process the event messages and take appropriate action. For example, certain events might require an operator to be informed, other events might start an application that performs some administration tasks automatically.

Event queues can have trigger actions associated with them and can create trigger messages. However, if these trigger messages in turn cause conditions that would normally generate an event, no event is generated. not generating an event in this instance ensures that looping does not occur.


Format of event messages

Event messages contain information about an event and its cause. Like other WebSphere MQ messages, an event message has two parts: a message descriptor and the message data.

Typically, you process event messages with a system management application tailored to meet the requirements of the enterprise at which it runs.

When the queue managers in a queue sharing group detect the conditions for generating an event message, several queue managers can generate an event message for the shared queue, resulting in several event messages. To ensure that a system can correlate multiple event messages from different queue managers, these event messages have a unique correlation identifier (CorrelId) set in the message descriptor (MQMD).


Related reference :

Activity report MQMD (message descriptor)

Activity report MQEPH (Embedded PCF header)

Activity report MQCFH (PCF header)


Related information :

Event message reference

Event message format

Event message MQMD (message descriptor)

Event message MQCFH (PCF header)

Event message descriptions


Concept Feedback


Last updated: Monday, September 23, 2013



Performance events

Performance events relate to conditions that can affect the performance of applications that use a specified queue. The scope of performance events is the queue. MQPUT calls and MQGET calls on one queue do not affect the generation of performance events on another queue.

Performance event messages can be generated at any appropriate time, not necessarily waiting until an MQI call for the queue is issued. However, if you use an MQI call on a queue to put or remove a message, any appropriate performance events are generated at that time.

Every performance event message that is generated is placed on the queue, SYSTEM.ADMIN.PERFM.EVENT.

The event data contains a reason code that identifies the cause of the event, a set of performance event statistics, and other data. The types of event data that can be returned in performance event messages are described in the following list:

Examples that illustrate the use of performance events assume that you set queue attributes by using the appropriate WebSphere MQ commands (MQSC). On , you can also set queue attributes using the operations and controls panels for queue managers.


Performance event statistics

The performance event data in the event message contains statistics about the event. Use the statistics to analyze the behavior of a specified queue.

The event data in the event message contains information about the event for system management programs. For all performance events, the event data contains the names of the queue manager and the queue associated with the event. The event data also contains statistics related to the event. Table 1 summarizes the event statistics that you can use to analyze the behavior of a queue. All the statistics refer to what has happened since the last time the statistics were reset.

Performance event statistics

Parameter Description
TimeSinceReset The elapsed time since the statistics were last reset.
HighQDepth The maximum number of messages on the queue since the statistics were last reset.
MsgEnqCount The number of messages enqueued (the number of MQPUT calls to the queue), since the statistics were last reset.
MsgDeqCount The number of messages dequeued (the number of MQGET calls to the queue), since the statistics were last reset.

Performance event statistics are reset when any of the following changes occur:


Queue service interval events

Queue service interval events indicate whether an operation was performed on a queue within a user-defined time interval called the service interval. Depending on your installation, you can use queue service interval events to monitor whether messages are being taken off queues quickly enough.

Queue service interval events are not supported on shared queues.

The following types of queue service interval events can occur, where the term get operation refers to an MQGET call or an activity that removes a messages from a queue, such as using the CLEAR QLOCAL command:

Queue Service Interval OK

Indicates that after one of the following operations:

  • An MQPUT call
  • A get operation that leaves a non-empty queue

a get operation was performed within a user-defined time period, known as the service interval.

Only a get operation can cause the Queue Service Interval OK event message. Queue Service Interval OK events are sometimes described as OK events.

Queue Service Interval High

Indicates that after one of the following operations:

  • An MQPUT call
  • A get operation that leaves a non-empty queue

a get operation was not performed within a user-defined service interval.

Either a get operation or an MQPUT call can cause the Queue Service Interval High event message. Queue Service Interval High events are sometimes described as High events.

To enable both Queue Service Interval OK and Queue Service Interval High events, set the QServiceIntervalEvent control attribute to High. Queue Service Interval OK events are automatically enabled when a Queue Service Interval High event is generated. You do not need to enable Queue Service Interval OK events independently.

OK and High events are mutually exclusive, so if one is enabled the other is disabled. However, both events can be simultaneously disabled.

Figure 1 shows a graph of queue depth against time. At time P1, an application issues an MQPUT, to put a message on the queue. At time G1, another application issues an MQGET to remove the message from the queue.

Figure 1. Understanding queue service interval events

The possible outcomes of queue service interval events are as follows:

The algorithm for starting the service timer and generating events is described in Rules for queue service interval events .


The service timer

Queue service interval events use an internal timer, called the service timer, which is controlled by the queue manager. The service timer is used only if a queue service interval event is enabled.

What precisely does the service timer measure?

The service timer measures the elapsed time between an MQPUT call to an empty queue or a get operation, and the next put or get, provided the queue depth is nonzero between these two operations.

When is the service timer active?

The service timer is always active (running), if the queue has messages on it (depth is nonzero) and a queue service interval event is enabled. If the queue becomes empty (queue depth zero), the timer is put into an OFF state, to be restarted on the next put.

When is the service timer reset?

The service timer is always reset after a get operation . It is also reset by an MQPUT call to an empty queue. However, it is not necessarily reset on a queue service interval event.

How is the service timer used?

Following a get operation or an MQPUT call, the queue manager compares the elapsed time as measured by the service timer, with the user-defined service interval. The result of this comparison is that:

  • An OK event is generated if there is a get operation and the elapsed time is less than or equal to the service interval, AND this event is enabled.
  • A high event is generated if the elapsed time is greater than the service interval, AND this event is enabled.

Can applications read the service timer?

No, the service timer is an internal timer that is not available to applications.

What about the TimeSinceReset parameter?

The TimeSinceReset parameter is returned as part of the event statistics in the event data. It specifies the time between successive queue service interval events, unless the event statistics are reset.


Concept Feedback


Last updated: Monday, September 23, 2013



Rules for queue service interval events

Formal rules control when the service timer is set and queue service interval events are generated.


Rules for the service timer

The service timer is reset to zero and restarted as follows:

The resetting of the timer does not depend on whether an event has been generated.

At queue manager startup the service timer is set to startup time if the queue depth is greater than zero.

If the queue is empty following a get operation, the timer is put into an OFF state.


Queue Service Interval High events

The Queue Service Interval event must be enabled (set to HIGH).

Queue Service Interval High events are automatically enabled when a Queue Service Interval OK event is generated.

If the service time is greater than the service interval, an event is generated on, or before, the next MQPUT or get operation.


Queue Service Interval OK events

Queue Service Interval OK events are automatically enabled when a Queue Service Interval High event is generated.

If the service time (elapsed time) is less than or equal to the service interval, an event is generated on, or before, the next get operation.


Related tasks :

Enabling queue service interval events


Concept Feedback


Last updated: Monday, September 23, 2013



Enabling queue service interval events

To configure a queue for queue service interval events you set the appropriate queue manager and queue attributes.

The high and OK events are mutually exclusive; that is, when one is enabled, the other is automatically disabled:

Enabling queue service interval events using MQSC

Queue service interval event Queue attributes

Queue Service Interval High
Queue Service Interval OK
No queue service interval events

QSVCIEV (HIGH)
QSVCIEV (OK)
QSVCIEV (NONE)

Service interval

QSVCINT (tt) where tt is the service
interval time in milliseconds.

Perform the following steps to enable queue service interval events:


Procedure

  1. Set the queue manager attribute PERFMEV to ENABLED. Performance events are enabled on the queue manager.
  2. Set the control attribute, QSVCIEV, for a Queue Service Interval High or OK event on the queue, as required.
  3. Set the QSVCINT attribute for the queue to specify the appropriate service interval time.


Example

To enable Queue Service Interval High events with a service interval time of 10 seconds (10 000 milliseconds) use the following MQSC commands:

  ALTER QMGR PERFMEV(ENABLED)

  ALTER QLOCAL('MYQUEUE') QSVCINT(10000) QSVCIEV(HIGH)
.


Task Feedback


Last updated: Monday, September 23, 2013



Queue service interval events examples

Use these examples to understand the information that you can obtain from queue service interval events

The three examples provide progressively more complex illustrations of the use of queue service interval events.

The figures accompanying the examples have the same structure:

The examples illustrate the following aspects of queue service interval events:

Remember: Example 1 shows a simple case where the messages are intermittent and each message is removed from the queue before the next one arrives. From the event data, you know that the maximum number of messages on the queue was one. You can, therefore, work out how long each message was on the queue.

However, in the general case, where there is more than one message on the queue and the sequence of MQGET calls and MQPUT calls is not predictable, you cannot use queue service interval events to calculate how long an individual message remains on a queue. The TimeSinceReset parameter, which is returned in the event data, can include a proportion of time when there are no messages on the queue. Therefore any results you derive from these statistics are implicitly averaged to include these times.


Queue service interval events: example 1

A basic sequence of MQGET calls and MQPUT calls, where the queue depth is always one or zero.

Figure 1. Queue service interval events - example 1


Commentary

  1. At P1, an application puts a message onto an empty queue. This starts the service timer.

    Note that T0 might be queue manager startup time.

  2. At G1, another application gets the message from the queue. Because the elapsed time between P1 and G1 is greater than the service interval, a Queue Service Interval High event is generated on the MQGET call at G1. When the high event is generated, the queue manager resets the event control attribute so that:

    1. The OK event is automatically enabled.
    2. The high event is disabled.

    Because the queue is now empty, the service timer is switched to an OFF state.

  3. At P2, a second message is put onto the queue. This restarts the service timer.
  4. At G2, the message is removed from the queue. However, because the elapsed time between P2 and G2 is less than the service interval, a Queue Service Interval OK event is generated on the MQGET call at G2. When the OK event is generated, the queue manager resets the control attribute so that:

    1. The high event is automatically enabled.
    2. The OK event is disabled.

    Because the queue is empty, the service timer is again switched to an OFF state.


Event statistics summary

Table 1 summarizes the event statistics for this example.

Event statistics summary for example 1

  Event 1 Event 2
Time of event T(G1) T(G2)
Type of event High OK
TimeSinceReset T(G1) - T(0) T(G2) - T(G1)
HighQDepth 1 1
MsgEnqCount 1 1
MsgDeqCount 1 1

The middle part of Figure 1 shows the elapsed time as measured by the service timer compared to the service interval for that queue. To see whether a queue service interval event might occur, compare the length of the horizontal line representing the service timer (with arrow) to that of the line representing the service interval. If the service timer line is longer, and the Queue Service Interval High event is enabled, a Queue Service Interval High event occurs on the next get. If the timer line is shorter, and the Queue Service Interval OK event is enabled, a Queue Service Interval OK event occurs on the next get.


Concept Feedback


Last updated: Monday, September 23, 2013



Queue service interval events: example 2

A sequence of MQPUT calls and MQGET calls, where the queue depth is not always one or zero.

This example also shows instances of the timer being reset without events being generated, for example, at time P2.

Figure 1. Queue service interval events - example 2


Commentary

In this example, OK events are enabled initially and queue statistics were reset at time T0.

  1. At P1, the first put starts the service timer.
  2. At P2, the second put does not generate an event because a put cannot cause an OK event.
  3. At G1, the service interval has now been exceeded and therefore an OK event is not generated. However, the MQGET call causes the service timer to be reset.
  4. At G2, the second get occurs within the service interval and this time an OK event is generated. The queue manager resets the event control attribute so that:

    1. The high event is automatically enabled.
    2. The OK event is disabled.

    Because the queue is now empty, the service timer is switched to an OFF state.


Event statistics summary

Table 1 summarizes the event statistics for this example.

Event statistics summary for example 2

  Event 2
Time of event T(G2)
Type of event OK
TimeSinceReset T(G2) - T(0)
HighQDepth 2
MsgEnqCount 2
MsgDeqCount 2


Concept Feedback


Last updated: Monday, September 23, 2013



Queue service interval events: example 3

A sequence of MQGET calls and MQPUT calls that is more sporadic than the previous examples.

Figure 1. Queue service interval events - example 3


Commentary

  1. At time T(0), the queue statistics are reset and Queue Service Interval High events are enabled.
  2. At P1, the first put starts the service timer.
  3. At P2, the second put increases the queue depth to two. A high event is not generated here because the service interval time has not been exceeded.
  4. At P3, the third put causes a high event to be generated. (The timer has exceeded the service interval.) The timer is not reset because the queue depth was not zero before the put. However, OK events are enabled.
  5. At G1, the MQGET call does not generate an event because the service interval has been exceeded and OK events are enabled. The MQGET call does, however, reset the service timer.
  6. At G2, the MQGET call does not generate an event because the service interval has been exceeded and OK events are enabled. Again, the MQGET call resets the service timer.
  7. At G3, the third get empties the queue and the service timer is equal to the service interval. Therefore an OK event is generated. The service timer is reset and high events are enabled. The MQGET call empties the queue, and this puts the timer in the OFF state.


Event statistics summary

Table 1 summarizes the event statistics for this example.

Event statistics summary for example 3

  Event 1 Event 2
Time of event T(P3) T(G3)
Type of event High OK
TimeSinceReset T(P3) - T(0) T(G3) - T(P3)
HighQDepth 3 3
MsgEnqCount 3 0
MsgDeqCount 0 3


Concept Feedback


Last updated: Monday, September 23, 2013



Queue depth events

Queue depth events are related to the queue depth, that is, the number of messages on the queue.

In WebSphere MQ applications, queues must not become full. If they do, applications can no longer put messages on the queue that they specify. Although the message is not lost if this occurs, a full queue can cause considerable inconvenience. The number of messages can build up on a queue if the messages are being put onto the queue faster than the applications that process them can take them off.

The solution to this problem depends on the particular circumstances, but might involve:

Advance warning that problems might be on their way makes it easier to take preventive action. For this purpose, WebSphere MQ provides the following queue depth events:

Queue Depth High events

Indicate that the queue depth has increased to a predefined threshold called the Queue Depth High limit.

Queue Depth Low events

Indicate that the queue depth has decreased to a predefined threshold called the Queue Depth Low limit.

Queue Full events

Indicate that the queue has reached its maximum depth, that is, the queue is full.

A Queue Full Event is generated when an application attempts to put a message on a queue that has reached its maximum depth. Queue Depth High events give advance warning that a queue is filling up. This means that having received this event, the system administrator needs to take some preventive action. You can configure the queue manager such that, if the preventive action is successful and the queue depth drops to a safer level, the queue manager generates a Queue Depth Low event.

The first queue depth event example illustrates the effect of presumed action preventing the queue becoming full.


Enabling queue depth events

To configure a queue for any of the queue depth events you set the appropriate queue manager and queue attributes. By default, all queue depth events are disabled. When enabled, queue depth events are generated as follows:

Perform the following steps to configure a queue for any of the queue depth events:


Procedure

  1. Enable performance events on the queue manager, using the queue manager attribute PERFMEV.

  2. Set one of the following attributes to enable the event on the required queue:
    • QDepthHighEvent (QDPHIEV in MQSC)
    • QDepthLowEvent (QDPLOEV in MQSC)
    • QDepthMaxEvent (QDPMAXEV in MQSC)

  3. Optional: To set the limits, assign the following attributes, as a percentage of the maximum queue depth:

    • QDepthHighLimit (QDEPTHHI in MQSC)
    • QDepthLowLimit (QDEPTHLO in MQSC)

    Restriction: QDEPTHHI must not be less than QDEPTHLO.

    If QDEPTHHI equals QDEPTHLO an event message is generated every time the queue depth passes the value in either direction, because the high threshold is enabled when the queue depth is below the value and the low threshold is enabled when the depth is above the value.

Note:

A Queue Depth Low event is not generated when expired messages are removed from a queue by a get operation causing the queue depth to be less than, or equal to, the value determined by the Queue Depth Low limit.

WebSphere MQ generates the low event message only during a successful get operation. Therefore, when the expired messages are removed from the queue, no queue depth low event message is generated.

Additionally, after the removal of these expired messages from the queue, queue depth high event and queue depth low event are not reset.


Example

To enable Queue Depth High events on the queue MYQUEUE with a limit set at 80%, use the following MQSC commands:

  ALTER QMGR PERFMEV(ENABLED)
  ALTER QLOCAL('MYQUEUE') QDEPTHHI(80) QDPHIEV(ENABLED)

To enable Queue Depth Low events on the queue MYQUEUE with a limit set at 20%, use the following MQSC commands:

  ALTER QMGR PERFMEV(ENABLED)
  ALTER QLOCAL('MYQUEUE') QDEPTHLO(20) QDPLOEV(ENABLED)

To enable Queue Full events on the queue MYQUEUE, use the following MQSC commands:

  ALTER QMGR PERFMEV(ENABLED)
  ALTER QLOCAL('MYQUEUE') QDPMAXEV(ENABLED)


Task Feedback


Last updated: Monday, September 23, 2013



Queue depth events examples

Use these examples to understand the information that you can obtain from queue depth events

The first example provides a basic illustration of queue depth events. The second example is more extensive, but the principles are the same as for the first example. Both examples use the same queue definition, as follows:

The queue, MYQUEUE1, has a maximum depth of 1000 messages. The high queue depth limit is 80% and the low queue depth limit is 20%. Initially, Queue Depth High events are enabled, while the other queue depth events are disabled.

The WebSphere MQ commands (MQSC) to configure this queue are:

  ALTER QMGR PERFMEV(ENABLED)
 
  DEFINE QLOCAL('MYQUEUE1') MAXDEPTH(1000) QDPMAXEV(DISABLED) QDEPTHHI(80)
    QDPHIEV(ENABLED) QDEPTHLO(20) QDPLOEV(DISABLED)


Queue depth events: example 1

A basic sequence of queue depth events.

Figure 1 shows the variation of queue depth over time.

Figure 1. Queue depth events (1)


Commentary

  1. At T(1), the queue depth is increasing (more MQPUT calls than MQGET calls) and crosses the Queue Depth Low limit. No event is generated at this time.
  2. The queue depth continues to increase until T(2), when the depth high limit (80%) is reached and a Queue Depth High event is generated.

    This enables both Queue Full and Queue Depth Low events.

  3. The (presumed) preventive actions instigated by the event prevent the queue from becoming full. By time T(3), the Queue Depth High limit has been reached again, this time from above. No event is generated at this time.
  4. The queue depth continues to fall until T(4), when it reaches the depth low limit (20%) and a Queue Depth Low event is generated.

    This enables both Queue Full and Queue Depth High events.


Event statistics summary

Table 1 summarizes the queue event statistics and Table 2 summarizes which events are enabled.

Event statistics summary for queue depth events (example 1)

  Event 2 Event 4
Time of event T(2) T(4)
Type of event Queue Depth High Queue Depth Low
TimeSinceReset T(2) - T(0) T(4) - T(2)
HighQDepth (Maximum queue depth since reset) 800 900
MsgEnqCount 1157 1220
MsgDeqCount 357 1820

Summary showing which events are enabled

Time period Queue Depth High event Queue Depth Low event Queue Full event
Before T(1) ENABLED - -
T(1) to T(2) ENABLED - -
T(2) to T(3) - ENABLED ENABLED
T(3) to T(4) - ENABLED ENABLED
After T(4) ENABLED - ENABLED


Concept Feedback


Last updated: Monday, September 23, 2013



Queue depth events: example 2

A more extensive sequence of queue depth events.

Figure 1 shows the variation of queue depth over time.

Figure 1. Queue depth events (2)


Commentary

  1. No Queue Depth Low event is generated at the following times:

    • T(1) (Queue depth increasing, and not enabled)
    • T(2) (Not enabled)
    • T(3) (Queue depth increasing, and not enabled)

  2. At T(4) a Queue Depth High event occurs. This enables both Queue Full and Queue Depth Low events.
  3. At T(9) a Queue Full event occurs after the first message that cannot be put on the queue because the queue is full.
  4. At T(12) a Queue Depth Low event occurs.


Event statistics summary

Table 1 summarizes the queue event statistics and Table 2 summarizes which events are enabled at different times for this example.

Event statistics summary for queue depth events (example 2)

  Event 4 Event 6 Event 8 Event 9 Event 12
Time of event T(4) T(6) T(8) T(9) T(12)
Type of event Queue Depth High Queue Depth Low Queue Depth High Queue Full Queue Depth Low
TimeSinceReset T(4) - T(0) T(6) - T(4) T(8) - T(6) T(9) - T(8) T(12) - T(9)
HighQDepth 800 855 800 1000 1000
MsgEnqCount 1645 311 1377 324 221
MsgDeqCount 845 911 777 124 1021

Summary showing which events are enabled

Time period Queue Depth High event Queue Depth Low event Queue Full event
T(0) to T(4) ENABLED - -
T(4) to T(6) - ENABLED ENABLED
T(6) to T(8) ENABLED - ENABLED
T(8) to T(9) - ENABLED ENABLED
T(9) to T(12) - ENABLED -
After T(12) ENABLED - ENABLED

Note: Events are out of syncpoint. Therefore you could have an empty queue, then fill it up causing an event, then roll back all of the messages under the control of a syncpoint manager. However, event enabling has been automatically set, so that the next time the queue fills up, no event is generated.


Concept Feedback


Last updated: Monday, September 23, 2013



Configuration events

Configuration events are notifications that are generated when an object is created, changed, or deleted, and can also be generated by explicit requests.

Configuration events notify you about changes to the attributes of an object. There are four types of configuration events:

The event data contains the following information:

Origin information

comprises the queue manager from where the change was made, the ID of the user that made the change, and how the change came about, for example by a console command.

Context information

a replica of the context information in the message data from the command message.

Context information is included in the event data only when the command was entered as a message on the SYSTEM.COMMAND.INPUT queue.

Object identity

comprises the name, type and disposition of the object.

Object attributes

comprises the values of all the attributes in the object.

In the case of change object events, two messages are generated, one with the information before the change, the other with the information after.

Every configuration event message that is generated is placed on the queue SYSTEM.ADMIN.CONFIG.EVENT.


Configuration event generation

Use this page to view the commands that cause configuration events to be generated and to understand the circumstances in which configuration events are not generated

A configuration event message is put to the configuration event queue when the CONFIGEV queue manager attribute is ENABLED and


When configuration events are not generated

Configuration events messages are not generated in the following circumstances:


Configuration event usage

Use this page to view how you can use configuration events to obtain information about your system, and to understand the factors, such as CMDSCOPE, that can affect your use of configuration events.

You can use configuration events for the following purposes:

  1. To produce and maintain a central configuration repository, from which reports can be produced and information about the structure of the system can be generated.
  2. To generate an audit trail. For example, if an object is changed unexpectedly, information regarding who made the alteration and when it was done can be stored.

    This can be particularly useful when command events are also enabled. If an MQSC or PCF command causes a configuration event and a command event to be generated, both event messages will share the same correlation identifier in their message descriptor.

For an MQSET call or any of the following commands:

if the queue manager attribute CONFIGEV is enabled, but the configuration event message cannot be put on the configuration event queue, for example the event queue has not been defined, the command or MQSET call is executed regardless.


Effects of CMDSCOPE

For commands where CMDSCOPE is used, the configuration event message or messages will be generated on the queue manager or queue managers where the command is executed, not where the command is entered. However, all the origin and context information in the event data will relate to the original command as entered, even where the command using CMDSCOPE is one that has been generated by the source queue manager.

Where a queue sharing group includes queue managers that are not at the current version, events will be generated for any command that is executed by means of CMDSCOPE on a queue manager that is at the current version, but not on those that are at a previous version. This happens even if the queue manager where the command is entered is at the previous version, although in such a case no context information is included in the event data.


Refresh Object configuration event

The Refresh Object configuration event is different from the other configuration events, because it occurs only when explicitly requested.

The create, change, and delete events are generated by an MQSET call or by a command to change an object but the refresh object event occurs only when explicitly requested by the MQSC command, REFRESH QMGR, or its PCF equivalent.

The REFRESH QMGR command is different from all the other commands that generate configuration events. All the other commands apply to a particular object and generate a single configuration event for that object. The REFRESH QMGR command can produce many configuration event messages potentially representing every object definition stored by a queue manager. One event message is generated for each object that is selected.

The REFRESH QMGR command uses a combination of three selection criteria to filter the number of objects involved:

If you specify none of the selection criteria on the REFRESH QMGR command, the default values are used for each selection criteria and a refresh configuration event message is generated for every object definition stored by the queue manager. This might cause unacceptable processing times and event message generation. Consider specifying some selection criteria.

The REFRESH QMGR command that generates the refresh events can be used in the following situations:


Command events

Command events are notifications that an MQSC, or PCF command has run successfully.

The event data contains the following information:

Origin information

comprises the queue manager from where the command was issued, the ID of the user that issued the command, and how the command was issued, for example by a console command.

Context information

a replica of the context information in the message data from the command message. If a command is not entered using a message, context information is omitted.

Context information is included in the event data only when the command was entered as a message on the SYSTEM.COMMAND.INPUT queue.

Command information

the type of command that was issued.

Command data

  • for PCF commands, a replica of the command data
  • for MQSC commands, the command text

The command data format does not necessarily match the format of the original command. For example, on distributed platforms the command data format is always in PCF format, even if the original request was an MQSC command.

Every command event message that is generated is placed on the command event queue, SYSTEM.ADMIN.COMMAND.EVENT.


Command event generation

Use this page to view the situations that cause command events to be generated and to understand the circumstances in which command events are not generated

A command event message is generated in the following situations:

If a command runs against the command event queue, SYSTEM.ADMIN.COMMAND.EVENT, a command event is generated if the queue still exists and it is not put-inhibited.


When command events are not generated

A command event message is not generated in the following circumstances:


Command event usage

Use this page to view how you can use command events to generate an audit trail of the commands that have run

For example, if an object is changed unexpectedly, information regarding who made the alteration and when it was done can be stored. This can be particularly useful when configuration events are also enabled. If an MQSC or PCF command causes a command event and a configuration event to be generated, both event messages will share the same correlation identifier in their message descriptor.

If a command event message is generated, but cannot be put on the command event queue, for example if the command event queue has not been defined, the command for which the command event was generated still runs regardless.


Effects of CMDSCOPE

For commands where CMDSCOPE is used, the command event message or messages will be generated on the queue manager or queue managers where the command runs, not where the command is entered. However, all the origin and context information in the event data will relate to the original command as entered, even where the command using CMDSCOPE is one that has been generated by the source queue manager.


Logger events

Logger events are notifications that a queue manager has started writing to a new log extent.

The event data contains the following information:

Every logger event message that is generated is placed on the logger event queue, SYSTEM.ADMIN.LOGGER.EVENT.


Logger event generation

Use this page to view the situations that cause logger events to be generated and to understand the circumstances in which logger events are not generated

A logger event message is generated in the following situations:

You can use the RESET QMGR MQSC command to request a queue manager to start writing to a new log extent.


When logger events are not generated

A logger event message is not generated in the following circumstances:


Logger event usage

Use this page to view how you can use logger events to determine the log extents that are no longer required for queue manager restart, or media recovery.

You can archive superfluous log extents to a medium such as tape for disaster recovery before removing them from the active log directory. Regular removal of superfluous log extents keeps disk space usage to a minimum.

If the LOGGEREV queue manager attribute is enabled, but a logger event message cannot be put on the logger event queue, for example because the event queue has not been defined, the action that caused the event continues regardless.


Sample program to monitor the logger event queue

Use this page to view a sample C program that monitors the logger event queue for new event messages, reads those messages, and puts the contents of the message to stdout.

 /******************************************************************************/
/*                                                                            */
/* Program name: AMQSLOG0.C                                                   */
/*                                                                            */
/* Description:  Sample C program to monitor the logger event queue and output*/
/*               a message to stdout when a logger event occurs               */
/* <N_OCO_COPYRIGHT>                                                          */
/* Licensed Materials - Property of IBM                                       */
/*                                                                            */
/* 63H9336                                                                    */
/* (c) Copyright IBM Corp. 2005 All Rights Reserved.                          */
/*                                                                            */
/* disclosure restricted by GSA ADP Schedule Contract with                    */
/* IBM Corp.                                                                  */
/* <NOC_COPYRIGHT>                                                            */
/******************************************************************************/
/*                                                                            */
/* Function: AMQSLOG is a sample program which monitors the logger event      */
/* queue for new event messages, reads those messages, and puts the contents  */
/* of the message to stdout.                                                  */
/*                                                                            */
/******************************************************************************/
/*                                                                            */
/* AMQSLOG has 1 parameter - the queue manager name (optional, if not         */
/* specified then the default queue manager is implied)                       */
/*                                                                            */
/******************************************************************************/

/******************************************************************************/
/* Includes                                                                   */
/******************************************************************************/
#include <stdlib.h>
#include <stdio.h>
#include <string.h>

#include <cmqc.h>        /* MQI constants*/
#include <cmqcfc.h>      /* PCF constants*/

/******************************************************************************/
/* Constants                                                                  */
/******************************************************************************/
                                     
#define   MAX_MESSAGE_LENGTH   8000

typedef struct _ParmTableEntry
{
  MQLONG  ConstVal;
  PMQCHAR Desc;
} ParmTableEntry;

ParmTableEntry ParmTable[] =
{
  0                             ,"",
  MQCA_Q_MGR_NAME               ,"Queue Manager Name",
  MQCMD_LOGGER_EVENT            ,"Logger Event Command",
  MQRC_LOGGER_STATUS            ,"Logger Status",
  MQCACF_CURRENT_LOG_EXTENT_NAME,"Current Log Extent",
  MQCACF_RESTART_LOG_EXTENT_NAME,"Restart Log Extent",
  MQCACF_MEDIA_LOG_EXTENT_NAME  ,"Media Log Extent",
  MQCACF_LOG_PATH               ,"Log Path"};              

/******************************************************************************/
/* Function prototypes                                                        */
/******************************************************************************/

static void ProcessPCF(MQHCONN    hConn,
                       MQHOBJ     hEventQueue,
                       PMQCHAR    pBuffer);

static PMQCHAR ParmToString(MQLONG Parameter);

/**********************************************************************/
/* Function: main                                                     */
/**********************************************************************/
int main(int argc, char * argv[])
{
  MQLONG    CompCode;
  MQLONG    Reason;
  MQHCONN   hConn = MQHC_UNUSABLE_HCONN;
  MQOD      ObjDesc = { MQOD_DEFAULT };
  MQCHAR    QMName[MQ_Q_MGR_NAME_LENGTH+1] = "";
  MQCHAR    LogEvQ[MQ_Q_NAME_LENGTH] = "SYSTEM.ADMIN.LOGGER.EVENT";
  MQHOBJ    hEventQueue;
  PMQCHAR   pBuffer = NULL;

  printf("\n/*************************************/\n");
  printf("/* Sample Logger Event Monitor start */\n");
  printf("/*************************************/\n");
  
  /********************************************************************/
  /* Parse any command line options                                   */
  /********************************************************************/
  
  if (argc > 1)
     strncpy(QMName, argv[1], (size_t)MQ_Q_MGR_NAME_LENGTH);

  pBuffer = (char *)malloc(MAX_MESSAGE_LENGTH);
  if (!pBuffer)
  {
    printf("Can't allocate %d bytes\n",MAX_MESSAGE_LENGTH);
    goto MOD_EXIT;
  }

  /********************************************************************/
  /* Connect to the specified (or default) queue manager              */
  /********************************************************************/

  MQCONN(QMName,
         &hConn,
         &CompCode,
         &Reason);

  if (Reason != MQCC_OK)
  {
    printf("Error in call to MQCONN, Reason %d, CompCode %d\n", Reason,
    CompCode);
    goto MOD_EXIT;
  }

  /* Open the logger event queue for input  */

  strncpy(ObjDesc.ObjectQMgrName,QMName, MQ_Q_MGR_NAME_LENGTH);
  strncpy(ObjDesc.ObjectName, LogEvQ, MQ_Q_NAME_LENGTH);

  MQOPEN(  hConn,
          &ObjDesc,
           MQOO_INPUT_EXCLUSIVE,
          &hEventQueue,
          &CompCode,
          &Reason);
  if (Reason)
  {
    printf("MQOPEN failed for queue manager %.48s Queue %.48s Reason: %d\n", 
				                              ObjDesc.ObjectQMgrName,
                                      ObjDesc.ObjectName,
                                      Reason);
    goto MOD_EXIT;
  }
  else
  {
    ProcessPCF(hConn, hEventQueue, pBuffer); 
  }

  MOD_EXIT:

  if (pBuffer != NULL) {
    free(pBuffer);
  }

  /********************************************************************/
  /* Disconnect                                                       */
  /********************************************************************/
  if (hConn != MQHC_UNUSABLE_HCONN) {
     MQDISC(&hConn, &CompCode, &Reason);
  }

  return 0;
}

/******************************************************************************/
/* Function: ProcessPCF                                                       */
/******************************************************************************/
/*                                                                            */
/* Input Parameters:  Handle to queue manager connection                      */
/*                    Handle to the opened logger event queue object          */
/*                    Pointer to a memory buffer to store the incoming PCF msg*/
/*                                                                            */
/* Output Parameters: None                                                    */
/*                                                                            */
/* Logic: Wait for messages to appear on the logger event queue and display   */
/* their contents.                                                            */
/*                                                                            */
/******************************************************************************/

static void ProcessPCF(MQHCONN    hConn,
                       MQHOBJ     hEventQueue,
                       PMQCHAR    pBuffer)
{
  MQCFH   * pCfh;
  MQCFST  * pCfst;
  MQGMO     Gmo     = { MQGMO_DEFAULT };
  MQMD      Mqmd    = { MQMD_DEFAULT };
  PMQCHAR   pPCFCmd;
  MQLONG    Reason  = 0;
  MQLONG    CompCode;
  MQLONG    MsgLen;
  PMQCHAR   Parm = NULL;
                                         /* Set timeout value           */
  Gmo.Options     |= MQGMO_WAIT;
  Gmo.Options |= MQGMO_CONVERT;
  Gmo.WaitInterval = MQWI_UNLIMITED;
  /********************************************************************/
  /* Process response Queue                                           */
  /********************************************************************/
  while (Reason == MQCC_OK)
  {
    memcpy(&Mqmd.MsgId;   , MQMI_NONE, sizeof(Mqmd.MsgId));
    memset(&Mqmd.CorrelId, 0, sizeof(Mqmd.CorrelId));

    MQGET( hConn,
           hEventQueue,
          &Mqmd,
          &Gmo,
           MAX_MESSAGE_LENGTH,
           pBuffer,
          &MsgLen,
          &CompCode,
          &Reason);
    if (Reason != MQCC_OK)
    {
      switch(Reason)
      {
        case MQRC_NO_MSG_AVAILABLE:
             printf("Timed out");
             break;

        default:
             printf("MQGET failed RC(%d)\n", Reason);
             break;
      }
      goto MOD_EXIT;
    }

    /******************************************************************/
    /* Only expect PCF event messages on this queue                   */
    /******************************************************************/
    if (memcmp(Mqmd.Format, MQFMT_EVENT, sizeof(Mqmd.Format)))
    {
     printf("Unexpected message format '%8.8s' received\n",Mqmd.Format);
     continue;
    }


    /*******************************************************************/
    /* Build the output by parsing the received PCF message, first the */
    /* header, then each of the parameters                             */
    /*******************************************************************/

    pCfh = (MQCFH *)pBuffer;

    if (pCfh -> Reason)
    {
     printf("-----------------------------------------------------------------\n");
     printf("Event Message Received\n");

     Parm = ParmToString(pCfh->Command);
     if (Parm != NULL) {
       printf("Command  :%s \n",Parm);
     }
     else
     {
       printf("Command  :%d \n",pCfh->Command);
     }
     
     printf("CompCode :%d\n"   ,pCfh->CompCode);

     Parm = ParmToString(pCfh->Reason);
     if (Parm != NULL) {
       printf("Reason   :%s \n",Parm);
     }
     else
     {
       printf("Reason   :%d \n",pCfh->Reason);
     }
    }

    pPCFCmd  = (char *)  (pCfh+1);
    printf("-----------------------------------------------------------------\n");
    while(pCfh -> ParameterCount--)
    {
      pCfst = (MQCFST *) pPCFCmd;
      switch(pCfst -> Type)
      {
        case MQCFT_STRING:
             Parm = ParmToString(pCfst -> Parameter);
             if (Parm != NULL) {
               printf("%-32s",Parm);
             }
             else
             {
               printf("%-32d",pCfst -> Parameter);
             }
             
             fwrite( pCfst -> String, pCfst -> StringLength, 1, stdout);
             pPCFCmd += pCfst -> StrucLength;
             break;
        default:
             printf("Unrecoginised datatype %d returned\n",pCfst->Type);
             goto MOD_EXIT;
      }
      putchar('\n');
    }
    printf("-----------------------------------------------------------------\n");
  }
MOD_EXIT:

 return;
}

/******************************************************************************/
/* Function: ParmToString                                                     */
/******************************************************************************/
/*                                                                            */
/* Input Parameters:  Parameter for which to get string description           */
/*                                                                            */
/* Output Parameters: None                                                    */
/*                                                                            */
/* Logic: Takes a parameter as input and returns a pointer to a string        */
/* description for that parameter, or NULL if the parameter does not  */
/* have an associated string description                                      */
/******************************************************************************/

static PMQCHAR ParmToString(MQLONG Parameter){
  long i;
  for (i=0 ; i< sizeof(ParmTable)/sizeof(ParmTableEntry); i++)
  {
    if (ParmTable[i].ConstVal == Parameter ParmTable[i].Desc)
      return ParmTable[i].Desc;
  }
  return NULL;
}


Sample output

This application produces the following form of output:

 /*************************************/                          
/* Sample Logger Event Monitor start */                          
/*************************************/                          
-----------------------------------------------------------------
Event Message Received                                           
Command  :Logger Event Command                                   
CompCode :0                                                      
Reason   :Logger Status                                          
-----------------------------------------------------------------
Queue Manager Name              CSIM                             
                                                                 
Current Log Extent              AMQA000001                       
Restart Log Extent              AMQA000001                       
Media Log Extent                AMQA000001                       
Log Path                        QMCSIM                           
-----------------------------------------------------------------


Sample program to monitor instrumentation events

Use this page to view a sample C program for monitoring instrumentation events

This sample program is not part of any WebSphere MQ product and is therefore not supplied as an actual physical item. The example is incomplete in that it does not enumerate all the possible outcomes of specified actions. However, you can use this sample as a basis for your own programs that use events, in particular, the PCF formats used in event messages. However, you need to modify this program before running it on your own systems.

 /********************************************************************/
 /*                                                                  */
 /* Program name: EVMON                                              */
 /*                                                                  */
 /* Description: C program that acts as an event monitor             */
 /*                                                                  */
 /*                                                                  */
 /********************************************************************/
 /*                                                                  */
 /* Function:                                                        */
 /*                                                                  */
 /*                                                                  */
 /*   EVMON is a C program that acts as an event monitor - reads an  */
 /*   event queue and tells you if anything appears on it            */
 /*                                                                  */
 /*   Its first parameter is the queue manager name, the second is   */
 /*   the event queue name. If these are not supplied it uses the    */
 /*   defaults.                                                      */
 /*                                                                  */
 /********************************************************************/
 #include <time.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
 #ifndef min
   #define min(a,b)        (((a) < (b)) ? (a) : (b))
 #endif

 /********************************************************************/
 /* includes for MQI                                                 */
 /********************************************************************/
 #include <cmqc.h>
 #include <cmqcfc.h>
 void printfmqcfst(MQCFST* pmqcfst);
 void printfmqcfin(MQCFIN* pmqcfst);
 void printreas(MQLONG reason);
 
  #define PRINTREAS(param)                                           \
     case param:                                                     \
       printf("Reason = %s\n",#param);                               \
       break;
 


 /********************************************************************/
 /* global variable                                                  */
 /********************************************************************/
 MQCFH    *evtmsg;                    /* evtmsg message buffer       */
 
 int main(int argc, char **argv)
 {
   /******************************************************************/
   /* declare variables                                              */
   /******************************************************************/
   int  i;                             /* auxiliary counter          */
   /******************************************************************/
   /* Declare MQI structures needed                                  */
   /******************************************************************/
   MQOD     od = {MQOD_DEFAULT};      /* Object Descriptor           */
   MQMD     md = {MQMD_DEFAULT};      /* Message Descriptor          */
   MQGMO   gmo = {MQGMO_DEFAULT};     /* get message options         */
   /******************************************************************/
   /* note, uses defaults where it can                               */
   /******************************************************************/

   MQHCONN  Hcon;                     /* connection handle           */
   MQHOBJ   Hobj;                     /* object handle               */
   MQLONG   O_options;                /* MQOPEN options              */
   MQLONG   C_options;                /* MQCLOSE options             */
   MQLONG   CompCode;                 /* completion code             */
   MQLONG   OpenCode;                 /* MQOPEN completion code      */
   MQLONG   Reason;                   /* reason code                 */
   MQLONG   CReason;                  /* reason code for MQCONN      */
   MQLONG   buflen;                   /* buffer length               */
   MQLONG   evtmsglen;                /* message length received     */
   MQCHAR   command[1100];            /* call command string ...     */
   MQCHAR   p1[600];                  /* ApplId insert               */
   MQCHAR   p2[900];                  /* evtmsg insert               */
   MQCHAR   p3[600];                  /* Environment insert          */
   MQLONG   mytype;                   /* saved application type      */
   char     QMName[50];               /* queue manager name          */
   MQCFST  *paras;                    /* the parameters              */
   int      counter;                  /* loop counter                */
   time_t   ltime;
 
   /******************************************************************/
   /* Connect to queue manager                                       */
   /******************************************************************/
   QMName[0] = 0;                      /* default queue manager     */
   if (argc > 1)
     strcpy(QMName, argv[1]);
   MQCONN(QMName,                      /* queue manager              */
          &Hcon,                   /* connection handle          */
          &CompCode,               /* completion code            */
          &CReason);               /* reason code                */
 


   /******************************************************************/
   /* Initialize object descriptor for subject queue                 */
   /******************************************************************/
   strcpy(od.ObjectName, "SYSTEM.ADMIN.QMGR.EVENT");
   if (argc > 2)
     strcpy(od.ObjectName, argv[2]);
 
   /******************************************************************/
   /* Open the event queue for input; exclusive or shared.  Use of   */
   /* the queue is controlled by the queue definition here           */
   /******************************************************************/

   O_options = MQOO_INPUT_AS_Q_DEF     /* open queue for input       */
         + MQOO_FAIL_IF_QUIESCING      /* but not if qmgr stopping   */
         + MQOO_BROWSE;
   MQOPEN(Hcon,                        /* connection handle          */
          &od,                     /* object descriptor for queue*/
          O_options,                   /* open options               */
          &Hobj,                   /* object handle              */
          &CompCode,               /* completion code            */
          &Reason);                /* reason code                */
 
   /******************************************************************/
   /*   Get messages from the message queue                          */
   /******************************************************************/
   while (CompCode != MQCC_FAILED)
   {
     /****************************************************************/
     /* I don't know how big this message is so just get the         */
     /* descriptor first                                             */
     /****************************************************************/
     gmo.Options = MQGMO_WAIT + MQGMO_LOCK
        + MQGMO_BROWSE_FIRST + MQGMO_ACCEPT_TRUNCATED_MSG;
                                       /* wait for new messages      */
     gmo.WaitInterval = MQWI_UNLIMITED;/* no time limit              */
     buflen = 0;                       /* amount of message to get   */
 
     /****************************************************************/
     /* clear selectors to get messages in sequence                  */
     /****************************************************************/
     memcpy(md.MsgId, MQMI_NONE, sizeof(md.MsgId));
     memcpy(md.CorrelId, MQCI_NONE, sizeof(md.CorrelId));
 
     /****************************************************************/
     /* wait for event message                                       */
     /****************************************************************/
     printf("...>\n");
     MQGET(Hcon,                       /* connection handle          */
           Hobj,                       /* object handle              */
           &md,                    /* message descriptor         */
           &gmo,                   /* get message options        */
           buflen,                     /* buffer length              */
           evtmsg,                     /* evtmsg message buffer      */
           &evtmsglen,             /* message length             */
           &CompCode,              /* completion code            */
           &Reason);               /* reason code                */
 


     /****************************************************************/
     /* report reason, if any                                        */
     /****************************************************************/
      if (Reason != MQRC_NONE && Reason != MQRC_TRUNCATED_MSG_ACCEPTED)
     {
       printf("MQGET ==> %ld\n", Reason);
     }
     else
     {
       gmo.Options = MQGMO_NO_WAIT + MQGMO_MSG_UNDER_CURSOR;
       buflen = evtmsglen;             /* amount of message to get   */
       evtmsg = malloc(buflen);
       if (evtmsg != NULL)
       {
         /************************************************************/
         /* clear selectors to get messages in sequence              */
         /************************************************************/
         memcpy(md.MsgId, MQMI_NONE, sizeof(md.MsgId));
         memcpy(md.CorrelId, MQCI_NONE, sizeof(md.CorrelId));
 
         /************************************************************/
         /* get the event message                                    */
         /************************************************************/
         printf("...>\n");
         MQGET(Hcon,                     /* connection handle        */
               Hobj,                     /* object handle            */
               &md,                  /* message descriptor       */
               &gmo,                 /* get message options      */
               buflen,                   /* buffer length            */
               evtmsg,                   /* evtmsg message buffer    */
               &evtmsglen,           /* message length           */
               &CompCode,            /* completion code          */
               &Reason);             /* reason code              */
 
         /************************************************************/
         /* report reason, if any                                    */
         /************************************************************/
         if (Reason != MQRC_NONE)
         {
           printf("MQGET ==> %ld\n", Reason);
         }
       }
       else
       {
         CompCode = MQCC_FAILED;
       }
     }


     /****************************************************************/
     /* . . . process each message received                          */
     /****************************************************************/

     if (CompCode != MQCC_FAILED)
     {
       /**************************************************************/
       /* announce a message                                         */
       /**************************************************************/
       printf("\a\a\a\a\a\a\a");
       time(&ltime);
       printf(ctime(&ltime));
 
       if (evtmsglen != buflen)
         printf("DataLength = %ld?\n", evtmsglen);
       else
       {
         /************************************************************/
         /* right let's look at the data                             */
         /************************************************************/
         if (evtmsg->Type != MQCFT_EVENT)
         {
           printf("Something's wrong this isn't an event message,"
                  " its type is %ld\n",evtmsg->Type);
         }
         else
         {
           if (evtmsg->Command == MQCMD_Q_MGR_EVENT)
           {
             printf("Queue Manager event: ");
           }
           else
             if (evtmsg->Command == MQCMD_CHANNEL_EVENT)
             {
               printf("Channel event: ");
             }
             else
        ⋮


             {
               printf("Unknown Event message, %ld.",
                       evtmsg->Command);
             }
 
           if      (evtmsg->CompCode == MQCC_OK)
             printf("CompCode(OK)\n");
           else if (evtmsg->CompCode == MQCC_WARNING)
             printf("CompCode(WARNING)\n");
           else if (evtmsg->CompCode == MQCC_FAILED)
             printf("CompCode(FAILED)\n");
           else
             printf("* CompCode wrong * (%ld)\n",
                       evtmsg->CompCode);
 
           if (evtmsg->StrucLength != MQCFH_STRUC_LENGTH)
           {
             printf("it's the wrong length, %ld\n",evtmsg->StrucLength);
           }
 
           if (evtmsg->Version != MQCFH_VERSION_1)
           {
             printf("it's the wrong version, %ld\n",evtmsg->Version);
           }
 
           if (evtmsg->MsgSeqNumber != 1)
           {
             printf("it's the wrong sequence number, %ld\n",
                     evtmsg->MsgSeqNumber);
           }
 
           if (evtmsg->Control != MQCFC_LAST)
           {
             printf("it's the wrong control option, %ld\n",
                     evtmsg->Control);
           }
 
           printreas(evtmsg->Reason);
           printf("parameter count is %ld\n", evtmsg->ParameterCount);
           /**********************************************************/
           /* get a pointer to the start of the parameters           */
           /**********************************************************/
 

           paras = (MQCFST *)(evtmsg + 1);
           counter = 1;
           while (counter <= evtmsg->ParameterCount)
           {
             switch (paras->Type)
             {
               case MQCFT_STRING:
                 printfmqcfst(paras);
                 paras = (MQCFST *)((char *)paras
                                      + paras->StrucLength);
                 break;
               case MQCFT_INTEGER:
                 printfmqcfin((MQCFIN*)paras);
                 paras = (MQCFST *)((char *)paras
                                      + paras->StrucLength);
                 break;
               default:
                 printf("unknown parameter type, %ld\n",
                        paras->Type);
                 counter = evtmsg->ParameterCount;
                 break;
             }
             counter++;
           }
         }
       }   /* end evtmsg action             */
       free(evtmsg);
       evtmsg = NULL;
     }     /* end process for successful GET */
   }       /* end message processing loop    */
 
   /******************************************************************/
   /* close the event queue - if it was opened                       */
   /******************************************************************/
   if (OpenCode != MQCC_FAILED)
   {
     C_options = 0;                  /* no close options             */
     MQCLOSE(Hcon,                   /* connection handle            */
            &Hobj,               /* object handle                */
            C_options,
            &CompCode,           /* completion code              */
            &Reason);            /* reason code                  */
   /******************************************************************/
   /* Disconnect from queue manager (unless previously connected)    */
   /******************************************************************/
   if (CReason != MQRC_ALREADY_CONNECTED)
   {
     MQDISC(&Hcon,               /* connection handle            */
            &CompCode,           /* completion code              */
            &Reason);            /* reason code                  */
 

 /********************************************************************/
 /*                                                                  */
 /* END OF EVMON                                                     */
 /*                                                                  */
 /********************************************************************/
 }
 
#define PRINTPARAM(param)                                          \
   case param:                                                     \
     {                                                             \
       char *p = #param;                                           \
     strncpy(thestring,pmqcfst->String,min(sizeof(thestring),      \
             pmqcfst->StringLength));                              \
     printf("%s %s\n",p,thestring);                                \
     }                                                             \
     break;
 
#define PRINTAT(param)                                             \
   case param:                                                     \
     printf("MQIA_APPL_TYPE = %s\n",#param);                       \
     break;
 
 
void printfmqcfst(MQCFST* pmqcfst)
{
  char thestring[100];
 
  switch (pmqcfst->Parameter)
  {
    PRINTPARAM(MQCA_BASE_Q_NAME)
    PRINTPARAM(MQCA_PROCESS_NAME)
    PRINTPARAM(MQCA_Q_MGR_NAME)
    PRINTPARAM(MQCA_Q_NAME)
    PRINTPARAM(MQCA_XMIT_Q_NAME)
    PRINTPARAM(MQCACF_APPL_NAME)
    ⋮
     default:
      printf("Invalid parameter, %ld\n",pmqcfst->Parameter);
      break;
  }
}
 


 
void printfmqcfin(MQCFIN* pmqcfst)
{
  switch (pmqcfst->Parameter)
  {
    case MQIA_APPL_TYPE:
      switch (pmqcfst->Value)
      {
        PRINTAT(MQAT_UNKNOWN)
        PRINTAT(MQAT_OS2)
        PRINTAT(MQAT_DOS)
        PRINTAT(MQAT_UNIX)
        PRINTAT(MQAT_QMGR)
        PRINTAT(MQAT_OS400)
        PRINTAT(MQAT_WINDOWS)
        PRINTAT(MQAT_CICS_VSE)
        PRINTAT(MQAT_VMS)
        PRINTAT(MQAT_GUARDIAN)
        PRINTAT(MQAT_VOS)
      }
      break;
    case MQIA_Q_TYPE:
      if (pmqcfst->Value == MQQT_ALIAS)
      {
        printf("MQIA_Q_TYPE is MQQT_ALIAS\n");
      }
      else
      ⋮
{
        if (pmqcfst->Value == MQQT_REMOTE)
        {
          printf("MQIA_Q_TYPE is MQQT_REMOTE\n");
          if (evtmsg->Reason == MQRC_ALIAS_BASE_Q_TYPE_ERROR)
          {
            printf("but remote is not valid here\n");
          }
        }
        else
        {
          printf("MQIA_Q_TYPE is wrong, %ld\n",pmqcfst->Value);
        }
      }
      break;
 

          case MQIACF_REASON_QUALIFIER:
      printf("MQIACF_REASON_QUALIFIER %ld\n",pmqcfst->Value);
      break;
 
    case MQIACF_ERROR_IDENTIFIER:
      printf("MQIACF_ERROR_INDENTIFIER %ld (X'%lX')\n",
              pmqcfst->Value,pmqcfst->Value);
      break;
 
    case MQIACF_AUX_ERROR_DATA_INT_1:
      printf("MQIACF_AUX_ERROR_DATA_INT_1 %ld (X'%lX')\n",
              pmqcfst->Value,pmqcfst->Value);
      break;
 
    case MQIACF_AUX_ERROR_DATA_INT_2:
      printf("MQIACF_AUX_ERROR_DATA_INT_2 %ld (X'%lX')\n",
              pmqcfst->Value,pmqcfst->Value);
      break;
⋮
default :
      printf("Invalid parameter, %ld\n",pmqcfst->Parameter);
      break;
  }
}
 
   void printreas(MQLONG reason)
{
  switch (reason)
  {
    PRINTREAS(MQRCCF_CFH_TYPE_ERROR)
    PRINTREAS(MQRCCF_CFH_LENGTH_ERROR)
    PRINTREAS(MQRCCF_CFH_VERSION_ERROR)
    PRINTREAS(MQRCCF_CFH_MSG_SEQ_NUMBER_ERR)
    ⋮
    PRINTREAS(MQRC_NO_MSG_LOCKED)
    PRINTREAS(MQRC_CONNECTION_NOT_AUTHORIZED)
    PRINTREAS(MQRC_MSG_TOO_BIG_FOR_CHANNEL)
    PRINTREAS(MQRC_CALL_IN_PROGRESS)
    default:
      printf("It's an unknown reason, %ld\n",
              reason);
      break;
  }
}
 
 


Message monitoring

Message monitoring is the process of identifying the route a message has taken through a queue manager network. By identifying the types of activities, and the sequence of activities performed on behalf of a message, the message route can be determined.

As a message passes through a queue manager network, various processes perform activities on behalf of the message. Use one of the following techniques to determine a message route:

These techniques all generate special messages that contain information about the activities performed on the message as it passed through a queue manager network. Use the information returned in these special messages to achieve the following objectives:


Activities and operations

Activities are discrete actions that an application performs on behalf of a message. Activities consist of operations, which are single pieces of work that an application performs.

The following actions are examples of activities:

Activities consist of one or more operations. Operations are single pieces of work that an application performs. For example, the activity of an MCA sending a message from a transmission queue down a channel consists of the following operations:

  1. Getting a message from a transmission queue (a Get operation).
  2. Sending the message down a channel (a Send operation).

In a publish/subscribe network, the activity of the WebSphere MQ publish/subscribe engine processing a message can consist of the following multiple operations:

  1. Putting a message to a topic string (a Put operation).
  2. Zero or more operations for each of the subscribers that are considered for receipt of the message (a Publish operation, a Discarded Publish operation or an Excluded Publish operation).


Information from activities

You can identify the sequence of activities performed on a message by recording information as the message is routed through a queue manager network. You can determine the route of a message through the queue manager network from the sequence of activities performed on the message, and can obtain the following information:

The last known location of a message

If a message does not reach its intended destination, you can determine the last known location of the message from a complete or partial message route.

Configuration issues with a queue manager network

When studying the route of a message through a queue manager network, you might see that the message has not gone where expected. There are many reasons why this can occur, for example, if a channel is inactive, the message might take an alternative route.

For a publish/subscribe application, you can also determine the route of a message being published to a topic and any messages that flow in a queue manager network as a result of being published to subscribers. In such situations, a system administrator can determine whether there are any problems in the queue manager network, and if appropriate, correct them.


Message routes

Depending on your reason for determining a message route, you can use the following general approaches:

Use activity information recorded for a trace-route message

Trace-route messages record activity information for a specific purpose. You can use them to determine configuration issues with a queue manager network, or to determine the last known location of a message. If a trace-route message is generated to determine the last known location of a message that did not reach its intended destination, it can mimic the original message. This gives the trace-route message the greatest chance of following the route taken by the original message.

The WebSphere MQ display route application can generate trace-route messages.

Use activity information recorded for the original message

You can enable any message for activity recording and have activity information recorded on its behalf. If a message does not reach its intended destination, you can use the recorded activity information to determine the last known location of the message. By using activity information from the original message, the most accurate possible message route can be determined, leading to the last known location. To use this approach, the original message must be enabled for activity recording.

Warning: Avoid enabling all messages in a queue manager network for activity recording. Messages enabled for activity recording can have many activity reports generated on their behalf. If every message in a queue manager network is enabled for activity recording, the queue manager network traffic can increase to an unacceptable level.


Message route techniques

Activity recording and trace-route messaging are techniques that allow you to record activity information for a message as it is routed through a queue manager network.

Activity recording

If a message has the appropriate report option specified, it requests that applications generate activity reports as it is routed through a queue manager network. When an application performs an activity on behalf of a message, an activity report can be generated, and delivered to an appropriate location. An activity report contains information about the activity that was performed on the message.

The activity information collected using activity reports must be arranged in order before a message route can be determined.

Trace-route messaging

Trace-route messaging is a technique that involves sending a trace-route message into a queue manager network. When an application performs an activity on behalf of the trace-route message, activity information can be accumulated in the message data of the trace-route message, or activity reports can be generated. If activity information is accumulated in the message data of the trace-route message, when it reaches its target queue a trace-route reply message containing all the information from the trace-route message can be generated and delivered to an appropriate location.

Because a trace-route message is dedicated to recording the sequence of activities performed on its behalf, there are more processing options available compared with normal messages that request activity reports.


Comparison of activity recording and trace-route messaging

Both activity recording and trace-route messaging can provide activity information to determine the route a message has taken through a queue manager network. Both methods have their own advantages.

Benefit Activity recording Trace-route messaging
Can determine the last known location of a message Yes Yes
Can determine configuration issues with a queue manager network Yes Yes

Can be requested by any message
(is not restricted to use with trace-route messages)

Yes No
Message data is left unmodified Yes No
Message processed normally Yes No
Activity information can be accumulated in the message data No Yes
Optional message delivery to target queue No Yes
If a message is caught in an infinite loop, it can be detected and dealt with No Yes
Activity information can be put in order reliably No Yes
Application provided to display the activity information No Yes


Message route completeness

In some cases it is not possible to identify the full sequence of activities performed on behalf of a message, so only a partial message route can be determined. The completeness of a message route is directly influenced by the queue manager network that the messages are routed through. The completeness of a message route depends on the level of the queue managers in the queue manager network, as follows:

Queue managers at WebSphere MQ Version 6.0 and subsequent releases

MCAs and user-written applications connected to queue managers at WebSphere MQ Version 6.0 or subsequent releases can record information related to the activities performed on behalf of a message. The recording of activity information is controlled by the queue manager attributes ACTIVREC and ROUTEREC. If a queue manager network consists of queue managers at WebSphere MQ Version 6.0 or subsequent releases only, complete message routes can be determined.

WebSphere MQ queue managers before Version 6.0

Applications connected to WebSphere MQ queue managers before Version 6.0 do not record the activities that they have performed on behalf of a message. If a queue manager network contains any WebSphere MQ queue manager prior to Version 6.0, only a partial message route can be determined.


How activity information is stored

WebSphere MQ stores activity information in activity reports, trace-route messages, or trace-route reply messages. In each case the information is stored in a structure called the Activity PCF group. A trace-route message or trace-route reply message can contain many Activity PCF groups, depending on the number of activities performed on the message. Activity reports contain one Activity PCF group because a separate activity report is generated for every recorded activity.

With trace-route messaging, additional information can be recorded. This additional information is stored in a structure called the TraceRoute PCF group. The TraceRoute PCF group contains a number of PCF structures that are used to store additional activity information, and to specify options that determine how the trace-route message is handled as it is routed through a queue manager network.


Activity recording

Activity recording is a technique for determining the routes that messages take through a queue manager network. To determine the route that a message has taken, the activities performed on behalf of the message are recorded.

When using activity recording, each activity performed on behalf of a message can be recorded in an activity report. An activity report is a type of report message. Each activity report contains information about the application that performed the activity on behalf of the message, when the activity took place, and information about the operations that were performed as part of the activity. Activity reports are typically delivered to a reply-to queue where they are collected together. By studying the activity reports related to a message, you can determine the route that the message took through the queue manager network.


Activity report usage

When messages are routed through a queue manager network, activity reports can be generated. You can use activity report information in the following ways:

Determine the last known location of a message

If a message that is enabled for activity recording does not reach its intended destination, activity reports generated for the message as it was routed through a queue manager network can be studied to determine the last known location of the message.

Determine configuration issues with a queue manager network

A number of messages enabled for activity recording can be sent into a queue manager network. By studying the activity reports related to each message it can become apparent that they have not taken the expected route. There are many reasons why this can occur, for example, a channel could have stopped, forcing the message to take an alternative route. In these situations, a system administrator can determine whether there are any problems in the queue manager network, and if there are, correct them.

Note: You can use activity recording in conjunction with trace-route messages by using the WebSphere MQ display route application.


Activity report format

Activity reports are PCF messages generated by applications that have performed an activity on behalf of a message. Activity reports are standard WebSphere MQ report messages containing a message descriptor and message data, as follows:

The message descriptor

  • An MQMD structure

Message data

  • An embedded PCF header (MQEPH)
  • Activity report message data

Activity report message data consists of the Activity PCF group, and if generated for a trace-route message, the TraceRoute PCF group.


Controlling activity recording

Enable activity recording at the queue manager level. To enable an entire queue manager network, individually enable every queue manager in the network for activity recording. If you enable more queue managers, more activity reports are generated.

To generate activity reports for a message as it is routed through a queue manager: define the message to request activity reports; enable the queue manager for activity recording; and ensure that applications performing activities on the message are capable of generating activity reports.

If you do not want activity reports to be generated for a message as it is routed through a queue manager, disable the queue manager for activity recording.


Procedure

  1. Request activity reports for a message

    1. In the message descriptor of the message, specify MQRO_ACTIVITY in the Report field.
    2. In the message descriptor of the message, specify the name of a reply-to queue in the ReplyToQ field.

    Warning: Avoid enabling all messages in a queue manager network for activity recording. Messages enabled for activity recording can have many activity reports generated on their behalf. If every message in a queue manager network is enabled for activity recording, the queue manager network traffic can increase to an unacceptable level.

  2. Enable or disable the queue manager for activity recording. Use the MQSC command ALTER QMGR, specifying the parameter ACTIVREC, to change the value of the queue manager attribute. The value can be:

    MSG

    The queue manager is enabled for activity recording. Any activity reports generated are delivered to the reply-to queue specified in the message descriptor of the message. This is the default value.

    QUEUE

    The queue manager is enabled for activity recording. Any activity reports generated are delivered to the local system queue SYSTEM.ADMIN.ACTIVITY.QUEUE. The system queue can also be used to forward activity reports to a common queue.

    DISABLED

    The queue manager is disabled for activity recording. No activity reports are generated while in the scope of this queue manager.
    For example, to enable a queue manager for activity recording and specify that any activity reports generated are delivered to the local system queue SYSTEM.ADMIN.ACTIVITY.QUEUE, use the following MQSC command:

     ALTER QMGR ACTIVREC(QUEUE)
    

    Remember: When you modify the ACTIVREC queue manager attribute, a running MCA does not detect the change until the channel is restarted.

  3. Ensure that your application uses the same algorithm as MCAs use to determine whether to generate an activity report for a message:

    1. Verify that the message has requested activity reports to be generated
    2. Verify that the queue manager where the message currently resides is enabled for activity recording
    3. Put the activity report on the queue determined by the ACTIVREC queue manager attribute


Task Feedback


Last updated: Monday, September 23, 2013



Set up a common queue for activity reports

To determine the locations of the activity reports related to a specific message when the reports are delivered to the local system queue, it is more efficient to use a common queue on a single node Set the ACTIVREC parameter to enable the queue manager for activity recording and to specify that any activity reports generated are delivered to the local system queue SYSTEM.ADMIN.ACTIVITY.QUEUE. If a number of queue managers in a queue manager network are set to deliver activity reports to the local system queue, it can be time consuming to determine the locations of the activity reports related to a specific message. Alternatively, use a single node, which is a queue manager that hosts a common queue. All the queue managers in a queue manager network can deliver activity reports to this common queue. The benefit of using a common queue is that queue managers do not have to deliver activity reports to the reply-to queue specified in a message and, when determining the locations of the activity reports related to a message, you query one queue only.

To set up a common queue, perform the following steps:


Procedure

  1. Select or define a queue manager as the single node

  2. On the single node, select or define a queue for use as the common queue

  3. On all queue managers where activity reports are to be delivered to the common queue, redefine the local system queue SYSTEM.ADMIN.ACTIVITY.QUEUE as a remote queue definition:

    1. Specify the name of the single node as the remote queue manager name
    2. Specify the name of the common queue as the remote queue name


Task Feedback


Last updated: Monday, September 23, 2013



Determining message route information

To determine a message route, obtain the information from the activity reports collected. Determine whether enough activity reports are on the reply-to queue to enable you to determine the required information and arrange the activity reports in order. The order that activity reports are put on the reply-to queue does not necessarily correlate to the order in which the activities were performed. You must order activity reports manually, unless they are generated for a trace-route message, in which case you can use the WebSphere MQ display route application to order the activity reports.

Determine whether enough activity reports are on the reply-to queue for you to obtain the necessary information:


Procedure

  1. Identify all related activity reports on the reply-to queue by comparing identifiers of the activity reports and the original message. Ensure you set the report option of the original message such that the activity reports can be correlated with the original message.

  2. Order the identified activity reports from the reply-to queue. You can use the following parameters from the activity report:

    OperationType

    The types of operations performed might enable you to determine the activity report that was generated directly before, or after, the current activity report.

    For example, an activity report details that an MCA sent a message from a transmission queue down a channel. The last operation detailed in the activity report has an OperationType of send and details that the message was sent using the channel, CH1, to the destination queue manager, QM1. This means that the next activity performed on the message will have occurred on queue manager, QM1, and that it will have begun with a receive operation from channel, CH1. By using this information you can identify the next activity report, providing it exists and has been acquired.

    OperationDate and OperationTime

    You can determine the general order of the activities from the dates and times of the operations in each activity report.

    Warning: Unless every queue manager in the queue manager network has their system clocks synchronized, ordering by date and time does not guarantee that the activity reports are in the correct sequence. You must establish the order manually.

    The order of the activity reports represents the route, or partial route, that the message took through the queue manager network.

  3. Obtain the information you need from the activity information in the ordered activity reports. If you have insufficient information about the message, you might be able to acquire further activity reports.


Task Feedback


Last updated: Monday, September 23, 2013



Retrieving further activity reports

To determine a message route, sufficient information must be available from the activity reports collected. If you retrieve the activity reports related to a message from the reply-to queue that the message specified, but you not have the necessary information, look for further activity reports. To determine the locations of any further activity reports, perform the following steps:


Procedure

  1. For any queue managers in the queue manager network that deliver activity reports to a common queue, retrieve activity reports from the common queue that have a CorrelId that matches the MsgId of the original message.

  2. For any queue managers in the queue manager network that do not deliver activity reports to a common queue, retrieve activity reports as follows:

    1. Examine the existing activity reports to identify queue managers through which the message was routed.
    2. For these queue managers, identify the queue managers that are enabled for activity recording.
    3. For these queue managers, identify any that did not return activity reports to the specified reply-to queue.
    4. For each of the queue managers that you identify, check the system queue SYSTEM.ADMIN.ACTIVITY.QUEUE and retrieve any activity reports that have a CorrelId that matches the MsgId of the original message.
    5. If you find no activity reports on the system queue, check the queue manager dead letter queue, if one exists. An activity report can only be delivered to a dead letter queue if the report option, MQRO_DEAD_LETTER_Q, is set.

  3. Arrange all the acquired activity reports in order. The order of the activity reports then represents the route, or partial route, that the message took.

  4. Obtain the information you need from the activity information in the ordered activity reports. In some circumstances, recorded activity information cannot reach the specified reply-to queue, a common queue, or a system queue.


Task Feedback


Last updated: Monday, September 23, 2013



Circumstances where activity information is not acquired

To determine the complete sequence of activities performed on behalf of a message, information related to every activity must be acquired. If the information relating to any activity has not been recorded, or has not been acquired, you can determine only a partial sequence of activities.

Activity information is not recorded in the following circumstances:

Recorded activity information is unable to reach the specified reply-to queue in the following circumstances:

Recorded activity information is unable to reach the system queue, or a common queue, in the following circumstances:

In these circumstances, providing the activity report does not have the report option MQRO_DISCARD_MSG specified, the activity report can be retrieved from a dead letter queue if one was defined on the queue manager where the activity report was rejected. An activity report will only have this report option specified if the original message, from which the activity report was generated, had both MQRO_PASS_DISCARD_AND_EXPIRY and MQRO_DISCARD_MSG specified in the Report field of the message descriptor.


Concept Feedback


Last updated: Monday, September 23, 2013



Trace-route messaging

Trace-route messaging is a technique that uses trace-route messages to record activity information for a message. Trace-route messaging involves sending a trace-route message into a queue manager network.

As the trace-route message is routed through the queue manager network, activity information is recorded. This activity information includes information about the applications that performed the activities, when they were performed, and the operations that were performed as part of the activities. You can use the information recorded using trace-route messaging for the following purposes:

To determine the last known location of a message

If a message does not reach its intended destination, you can use the activity information recorded for a trace-route message to determine the last known location of the message. A trace-route message is sent into a queue manager network with the same target destination as the original message, intending that it follows the same route. Activity information can be accumulated in the message data of the trace-route message, or recorded using activity reports. To increase the probability that the trace-route message follows the same route as the original message, you can modify the trace-route message to mimic the original message.

To determine configuration issues with a queue manager network

Trace-route messages are sent into a queue manager network and activity information is recorded. By studying the activity information recorded for a trace-route message, it can become apparent that the trace-route message did not follow the expected route. There are many reasons why this can occur, for example, a channel might be inactive, forcing the message to take an alternative route. In these situations, a system administrator can determine whether there are any problems in the queue manager network, and if there are, correct them.

You can use the WebSphere MQ display route application to configure, generate, and put trace-route messages into a queue manager network.

Warning: If you put a trace-route message to a distribution list, the results are undefined.


How activity information is recorded

With trace-route messaging, you can record activity information in the message data of the trace-route message, or use activity reports. Alternatively, you can use both techniques.


Accumulating activity information in the message data of the trace-route message

As a trace-route message is routed through a queue manager network, information about the activities performed on behalf of the trace-route message can be accumulated in the message data of the trace-route message. The activity information is stored in Activity PCF groups. For every activity performed on behalf of the trace-route message, an Activity PCF group is written to the end of the PCF block in the message data of the trace-route message.

Additional activity information is recorded in trace-route messaging, in a PCF group called the TraceRoute PCF group. The additional activity information is stored in this PCF group, and can be used to help determine the sequence of recorded activities. This technique is controlled by the Accumulate parameter in the TraceRoute PCF group.


Recording activity information using activity reports

As a trace-route message is routed through a queue manager network, an activity report can be generated for every activity that was performed on behalf of the trace-route message. The activity information is stored in the Activity PCF group. For every activity performed on behalf of a trace-route message, an activity report is generated containing an Activity PCF group. Activity recording for trace-route messages works in the same way as for any other message.

Activity reports generated for trace-route messages contain additional activity information compared to the those generated for any other message. The additional information is returned in a TraceRoute PCF group. The information contained in the TraceRoute PCF group is accurate only from the time the activity report was generated. You can use the additional information to help determine the sequence of activities performed on behalf of the trace-route message.


Concept Feedback


Last updated: Monday, September 23, 2013



Acquiring recorded activity information

When a trace-route message has reached its intended destination, or is discarded, the method that you use to acquire the activity information depends on how that information was recorded. If you are unfamiliar with activity information, refer to How activity information is recorded . Use the following methods to acquire the activity information after the trace-route message has reached its intended destination, or is discarded:


Procedure


Task Feedback


Last updated: Monday, September 23, 2013



Controlling trace-route messaging

Enable trace-route messaging at the queue manager level, so that applications in the scope of that queue manager can write activity information to a trace-route message. To enable an entire queue manager network, individually enable every queue manager in the network for trace-route messaging. If you enable more queue managers, more activity reports are generated. If you are using activity reports to record activity information for a trace-route message, refer to Controlling activity recording . To record activity information for a trace-route message as it is routed through a queue manager, perform the following steps:


Procedure


Enabling queue managers for trace-route messaging

To control whether queue managers are enabled or disabled for trace-route messaging use the queue manager attribute ROUTEREC.

Use the MQSC command ALTER QMGR, specifying the parameter ROUTEREC to change the value of the queue manager attribute. The value can be:

MSG

The queue manager is enabled for trace-route messaging. Applications within the scope of the queue manager can write activity information to the trace-route message.

If the Accumulate parameter in the TraceRoute PCF group is set as MQROUTE_ACCUMULATE_AND_REPLY, and the next activity to be performed on the trace-route message:

  • is a discard
  • is a put to a local queue (target queue or dead-letter queue)
  • will cause the total number of activities performed on the trace-route message to exceed the value of parameter the MaxActivities, in the TraceRoute PCF group .

a trace-route reply message is generated, and delivered to the reply-to queue specified in the message descriptor of the trace-route message.

QUEUE

The queue manager is enabled for trace-route messaging. Applications within the scope of the queue manager can write activity information to the trace-route message.

If the Accumulate parameter in the TraceRoute PCF group is set as MQROUTE_ACCUMULATE_AND_REPLY, and the next activity to be performed on the trace-route message:

  • is a discard
  • is a put to a local queue (target queue or dead-letter queue)
  • will cause the total number of activities performed on the trace-route message to exceed the value of parameter the MaxActivities, in the TraceRoute PCF group .

a trace-route reply message is generated, and delivered to the local system queue SYSTEM.ADMIN.TRACE.ROUTE.QUEUE.

DISABLED

The queue manager is disabled for trace-route messaging. Activity information is not accumulated in the the trace-route message, however the TraceRoute PCF group can be updated while in the scope of this queue manager.

For example, to disable a queue manager for trace-route messaging, use the following MQSC command:

 ALTER QMGR ROUTEREC(DISABLED)

Remember: When you modify the ROUTEREC queue manager attribute, a running MCA does not detect the change until the channel is restarted.


Reference Feedback


Last updated: Monday, September 23, 2013



Enabling applications for trace-route messaging

To enable trace-route messaging for a user application, base your algorithm on the algorithm used by message channel agents (MCAs) If you are not familiar with the format of a trace-route message, see Trace-route message reference . Message channel agents (MCAs) are enabled for trace-route messaging. To enable a user application for trace-route messaging, use the following steps from the algorithm that MCAs use:


Procedure

  1. Determine whether the message being processed is a trace-route message. If the message does not conform to the format of a trace-route message, the message is not processed as a trace-route message.

  2. Determine whether activity information is to be recorded. If the detail level of the performed activity is not less than the level of detail specified by the Detail parameter, activity information is recorded under specific circumstances. This information is only recorded if the trace-route message requests accumulation, and the queue manager is enabled for trace-route messaging, or if the trace-route message requests an activity report and the queue manager is enabled for activity recording.

    • If activity information is to be recorded, increment the RecordedActivities parameter.
    • If activity information is not to be recorded, increment the UnrecordedActivities parameter.

  3. Determine whether the total number of activities performed on the trace-route message exceeds the value of the MaxActivities parameter.

    The total number of activities is the sum of RecordedActivities, UnrecordedActivities, and DiscontinuityCount.

    If the total number of activities exceeds MaxActivities, reject the message with feedback MQFB_MAX_ACTIVITIES.

  4. If value of Accumulate is set as MQROUTE_ACCUMULATE_IN_MSG or MQROUTE_ACCUMULATE_AND_REPLY, and the queue manager is enabled for trace-route messaging, write an Activity PCF group to the end of the PCF block in the message data of a trace-route message.

  5. Deliver the trace-route message to a local queue.
    • If the parameter, Deliver, is specified as MQROUTE_DELIVER_NO, reject the trace-route message with feedback MQFB_NOT_DELIVERED.
    • If the parameter, Deliver, is specified as MQROUTE_DELIVER_YES, deliver the trace-route message to the local queue.

  6. Generate a trace-route reply message if all the following conditions are true:

    • The trace-route message was delivered to a local queue or rejected
    • The value of the parameter, Accumulate, is MQROUTE_ACCUMULATE_AND_REPLY
    • The queue manager is enabled for trace-route messaging

    The trace-route reply message is put on the queue determined by the ROUTEREC queue manager attribute.

  7. If the trace-route message requested an activity report and the queue manager is enabled for activity recording, generate an activity report. The activity report is put on the queue determined by the ACTIVREC queue manager attribute.


Task Feedback


Last updated: Monday, September 23, 2013



Generating and configuring a trace-route message

A trace-route message comprises specific message descriptor and message data parts. To generate a trace-route message, either create the message manually or use the WebSphere MQ display route application.

A trace-route message consists of the following parts:

Message descriptor

An MQMD structure, with the Format field set to MQFMT_ADMIN or MQFMT_EMBEDDED_PCF.

Message data

One of the following combinations:

  • A PCF header (MQCFH) and trace-route message data, if Format is set to MQFMT_ADMIN
  • An embedded PCF header (MQEPH), trace-route message data, and additional user-specified message data, if Format is set to MQFMT_EMBEDDED_PCF

The trace-route message data consists of the TraceRoute PCF group and one or more Activity PCF groups.


Manual generation

When generating a trace-route message manually, an Activity PCF group is not required. Activity PCF groups are written to the message data of the trace-route message when an MCA or user-written application performs an activity on its behalf.


The WebSphere MQ display route application

Use the WebSphere MQ display route application, dspmqrte, to configure, generate and put a trace-route message into a queue manager network. Set the Format parameter in the message descriptor to MQFMT_ADMIN. You cannot add user data to the trace-route message generated by the WebSphere MQ display route application.

Restriction: dspmqrte cannot be issued on queue managers before WebSphere MQ Version 6.0 or on WebSphere MQ for z/OS queue managers. The first queue manager the trace-route message is routed through to be a queue manager of this type, connect to the queue manager as a WebSphere MQ Version 6.0 or later client using the optional parameter -c.


Mimicking the original message

When using a trace-route message to determine the route another message has taken through a queue manager network, the more closely a trace-route message mimics the original message, the greater the chance that the trace-route message will follow the same route as the original message.

The following message characteristics can affect where a message is forwarded to within a queue manager network:

Priority

The priority can be specified in the message descriptor of the message.

Persistence

The persistence can be specified in the message descriptor of the message.

Expiration

The expiration can be specified in the message descriptor of the message.

Report options

Report options can be specified in the message descriptor of the message.

Message size

To mimic the size of a message, additional data can be written to the message data of the message. For this purpose, additional message data can be meaningless.

The WebSphere MQ display route application cannot specify message size.

Message data

Some queue manager networks use content based routing to determine where messages are forwarded. In these cases the message data of the trace-route message needs to be written to mimic the message data of the original message.

The WebSphere MQ display route application cannot specify message data.


Reference Feedback


Last updated: Monday, September 23, 2013



The TraceRoute PCF group

Attributes in the TraceRoute PCF group control the behavior of a trace-route message. The TraceRoute PCF group is in the message data of every trace-route message.

The following table lists the parameters in the TraceRoute group that an MCA recognizes. Further parameters can be added if user-written applications are written to recognize them, as described in Additional activity information .

TraceRoute PCF group

Parameter Type

TraceRoute
   Detail
   RecordedActivities
   UnrecordedActivities
   DiscontinuityCount
   MaxActivities
   Accumulate
   Forward
   Deliver

MQCFGR
MQCFIN
MQCFIN
MQCFIN
MQCFIN
MQCFIN
MQCFIN
MQCFIN
MQCFIN

Descriptions of each parameter in the TraceRoute PCF group follows:

Detail

Specifies the detail level of activity information that is to be recorded. The value can be:

MQROUTE_DETAIL_LOW

Only activities performed by user application are recorded.

MQROUTE_DETAIL_MEDIUM

Activities specified in MQROUTE_DETAIL_LOW should be recorded. Additionally, activities performed by MCAs are recorded.

MQROUTE_DETAIL_HIGH

Activities specified in MQROUTE_DETAIL_LOW, and MQROUTE_DETAIL_MEDIUM should be recorded. MCAs do not record any further activity information at this level of detail. This option is only available to user applications that are to record further activity information. For example, if a user application determines the route a message takes by considering certain message characteristics, the information about the routing logic could be included with this level of detail.

RecordedActivities

Specifies the number of recorded activities performed on behalf of the trace-route message. An activity is considered to be recorded if information about it has been written to the trace-route message, or if an activity report has been generated. For every recorded activity, RecordedActivities increments by one.

UnrecordedActivities

Specifies the number of unrecorded activities performed on behalf of the trace-route message. An activity is considered to be unrecorded if an application that is enabled for trace-route messaging neither accumulates, nor writes the related activity information to an activity report.

An activity performed on behalf of a trace-route message is unrecorded in the following circumstances:

  • The detail level of the performed activity is less than the level of detail specified by the parameter Detail.
  • The trace-route message requests an activity report but not accumulation, and the queue manager is not enabled for activity recording.
  • The trace-route message requests accumulation but not an activity report, and the queue manager is not enabled for trace-route messaging.
  • The trace-route message requests both accumulation and an activity report, and the queue manager is not enabled for activity recording and trace route messaging.
  • The trace-route message requests neither accumulation nor an activity report.

For every unrecorded activity the parameter, UnrecordedActivities, increments by one.

DiscontinuityCount

Specifies the number of times the trace-route message has been routed through a queue manager with applications that were not enabled for trace-route messaging. This value is incremented by the queue manager. If this value is greater than 0, only a partial message route can be determined.

MaxActivities

Specifies the maximum number of activities that can be performed on behalf of the trace-route message.

The total number of activities is the sum of RecordedActivities, UnrecordedActivities, and DiscontinuityCount. The total number of activities must not exceed the value of MaxActivities.

The value of MaxActivities can be:

A positive integer

The maximum number of activities.

If the maximum number of activities is exceeded, the trace-route message is rejected with feedback MQFB_MAX_ACTIVITIES. This can prevent the trace-route message from being forwarded indefinitely if caught in an infinite loop.

MQROUTE_UNLIMITED_ACTIVITIES

An unlimited number of activities can be performed on behalf of the trace-route message.

Accumulate

Specifies the method used to accumulate activity information. The value can be:

MQROUTE_ACCUMULATE_IN_MSG

If the queue manager is enabled for trace-route messaging, activity information is accumulated in the message data of the trace-route message.

If this value is specified, the trace-route message data consists of the following:

  • The TraceRoute PCF group.
  • Zero or more Activity PCF groups.

MQROUTE_ACCUMULATE_AND_REPLY

If the queue manager is enabled for trace-route messaging, activity information is accumulated in the message data of the trace-route message, and a trace-route reply message is generated if any of the following occur:

  • The trace-route message is discarded by a WebSphere MQ Version 6 (or later) queue manager.
  • The trace-route message is put to a local queue (target queue or dead-letter queue) by a WebSphere MQ Version 6 (or later) queue manager.
  • The number of activities performed on the trace-route message exceeds the value of MaxActivities.

If this value is specified, the trace-route message data consists of the following:

  • The TraceRoute PCF group.
  • Zero or more Activity PCF groups.

MQROUTE_ACCUMULATE_NONE

Activity information is not accumulated in the message data of the trace-route message.

If this value is specified, the trace-route message data consists of the following:

  • The TraceRoute PCF group.

Forward

Specifies where a trace-route message can be forwarded to. The value can be:

MQROUTE_FORWARD_IF_SUPPORTED

The trace-route message is only forwarded to queue managers that will honor the value of the Deliver parameter from the TraceRoute group.

MQROUTE_FORWARD_ALL

The trace-route message is forwarded to any queue manager, regardless of whether the value of the Deliver parameter will be honored.
Queue managers use the following algorithm when determining whether to forward a trace-route message to a remote queue manager:

  1. Determine whether the remote queue manager is capable of supporting trace-route messaging.

    • If the remote queue manager is capable of supporting trace-route messaging, the algorithm continues to step 4 .
    • If the remote queue manager is not capable of supporting trace-route messaging, the algorithm continues to step 2

  2. Determine whether the Deliver parameter from the TraceRoute group contains any unrecognized delivery options in the MQROUTE_DELIVER_REJ_UNSUP_MASK bit mask.

    • If any unrecognized delivery options are found, the trace-route message is rejected with feedback MQFB_UNSUPPORTED_DELIVERY.
    • If no unrecognized delivery options are found, the algorithm continues to step 3 .

  3. Determine the value of the parameter Deliver from the TraceRoute PCF group in the trace-route message.

    • If Deliver is specified as MQROUTE_DELIVER_YES, the trace-route message is forwarded to the remote queue manager.
    • If Deliver is specified as MQROUTE_DELIVER_NO, the algorithm continues to step 4 .

  4. Determine whether the Forward parameter from the TraceRoute group contains any unrecognized forwarding options in the MQROUTE_FORWARDING_REJ_UNSUP_MASK bit mask.

    • If any unrecognized forwarding options are found, the trace-route message is rejected with feedback MQFB_UNSUPPORTED_FORWARDING.
    • If no unrecognized forwarding options are found, the algorithm continues to step 5 .

  5. Determine the value of the parameter Forward from the TraceRoute PCF group in the trace-route message.

    • If Forward is specified as MQROUTE_FORWARD_IF_SUPPORTED, the trace-route message is rejected with feedback MQFB_NOT_FORWARDED.
    • If Forward is specified as MQROUTE_FORWARD_ALL, trace-route message can be forwarded to the remote queue manager.

Deliver

Specifies the action to be taken if the trace-route message reaches its intended destination. User-written applications must check this attribute before placing a trace-route message on its target queue. The value can be:

MQROUTE_DELIVER_YES

On arrival, the trace-route message is put on the target queue. Any application performing a get operation on the target queue can retrieve the trace-route message.

MQROUTE_DELIVER_NO

On arrival, the trace-route message is not delivered to the target queue. The message is processed according to its report options.


Reference Feedback


Last updated: Monday, September 23, 2013



Set up a common queue for trace-route reply messages

To determine the locations of the trace-route reply messages related to a specific message when the reports are delivered to the local system queue, it is more efficient to use a common queue on a single node Set the ROUTEREC parameter to enable the queue manager for trace-route messaging and to specify that any trace-route reply messages generated are delivered to the local system queue SYSTEM.ADMIN.TRACE.ROUTE.QUEUE. If a number of queue managers in a queue manager network are set to deliver trace-route reply messages to the local system queue, it can be time consuming to determine the locations of the trace-route reply messages related to a specific message. Alternatively, use a single node, which is a queue manager that hosts a common queue. All the queue managers in a queue manager network can deliver trace-route reply messages to this common queue. The benefit of using a common queue is that queue managers do not have to deliver trace-route reply messages to the reply-to queue specified in a message and, when determining the locations of the trace-route reply messages related to a message, you query one queue only.

To set up a common queue, perform the following steps:


Procedure

  1. Select or define a queue manager as the single node

  2. On the single node, select or define a queue for use as the common queue

  3. On all queue managers that forward trace-route reply messages to the common queue, redefine the local system queue SYSTEM.ADMIN.TRACE.ROUTE.QUEUE as a remote queue definition

    1. Specify the name of the single node as the remote queue manager name
    2. Specify the name of the common queue as the remote queue name


Task Feedback


Last updated: Monday, September 23, 2013



Acquiring and using recorded information

Use any of the following techniques to acquire recorded activity information for a trace-route message

Note that the circumstances in which activity information is not acquired apply also to trace-route reply messages.

Activity information is not recorded when a trace-route message is processed by a queue manager that is disabled for both activity recording and trace-route messaging.


Acquiring information from trace-route reply messages

To acquire activity information you locate the trace-route reply message. Then you retrieve the message and analyze the activity information. You can acquire activity information from a trace-route reply message only if you know the location of the trace-route reply message. Locate the message and process the activity information as follows:


Procedure

  1. Check the reply-to queue that was specified in the message descriptor of the trace-route message. If the trace-route reply message is not on the reply-to queue, check the following locations:
    • The local system queue, SYSTEM.ADMIN.TRACE.ROUTE.QUEUE, on the target queue manager of the trace-route message
    • The common queue, if you have set up a common queue for trace-route reply messages
    • The local system queue, SYSTEM.ADMIN.TRACE.ROUTE.QUEUE, on any other queue manager in the queue manager network, which can occur if the trace-route message has been put to a dead-letter queue, or the maximum number of activities was exceeded

  2. Retrieve the trace-route reply message

  3. Use the WebSphere MQ display route application to display the recorded activity information

  4. Study the activity information and obtain the information that you need


Task Feedback


Last updated: Monday, September 23, 2013



Acquiring information from trace-route messages

To acquire activity information you locate the trace-route message, which must have the appropriate parameters in the TraceRoute PCF group. Then you retrieve the message and analyze the activity information. You can acquire activity information from a trace-route message only if you know the location of the trace-route message and it has the parameter Accumulate in the TraceRoute PCF group specified as either MQROUTE_ACCUMULATE_IN_MSG or MQROUTE_ACCUMULATE_AND_REPLY.

For the trace-route message to be delivered to the target queue the Deliver parameter in the TraceRoute PCF group must be specified as MQROUTE_DELIVER_YES.


Procedure

  1. Check the target queue. If the trace-route message is not on the target queue, you can try to locate the trace-route message using a trace-route message enabled for activity recording. With the generated activity reports try to determine the last known location of the trace-route message.
  2. Retrieve the trace-route message
  3. Use the WebSphere MQ display route application to display the recorded activity information
  4. Study the activity information and obtain the information that you need


Task Feedback


Last updated: Monday, September 23, 2013



Acquiring information from activity reports

To acquire activity information you locate the activity report, which must have the report option specified in the message descriptor. Then you retrieve the activity report and analyze the activity information. You can acquire activity information from an activity report only if you know the location of the activity report and the report option MQRO_ACTIVITY was specified in the message descriptor of the trace-route message.


Procedure

  1. Locate and order the activity reports generated for a trace-route message. When you have located the activity reports, you can order them manually or use the WebSphere MQ display route application to order and display the activity information automatically.
  2. Study the activity information and obtain the information that you need


Task Feedback


Last updated: Monday, September 23, 2013



Additional activity information

As a trace-route message is routed through a queue manager network, user applications can record additional information by including one or more additional PCF parameters when writing the Activity group to the message data of the trace-route message or activity report.

Additional activity information can help system administrators to identify the route taken by a trace-route message took, or why that route was taken.

If you use the WebSphere MQ display route application to display the recorded information for a trace-route message, any additional PCF parameters can only be displayed with a numeric identifier, unless the parameter identifier of each parameter is recognized by the WebSphere MQ display route application. To recognize a parameter identifier, additional information must be recorded using the following PCF parameters. Include these PCF parameters in an appropriate place in the Activity PCF group.

GroupName

Group name

Description Grouped parameters specifying the additional information.
Identifier MQGACF_VALUE_NAMING.
Data type MQCFGR
Parameters in group
  • ParameterName
  • ParameterValue

ParameterName

Parameter name

Description Contains the name to be displayed by the WebSphere MQ display route application, which puts the value of ParameterValue into context.
Identifier MQCA_VALUE_NAME.
Data type MQCFST
Included in PCF group: GroupName.
Value: The name to be displayed.

ParameterValue

Parameter value

Description Contains the value to be displayed by the WebSphere MQ display route application.
Identifier: The PCF structure identifier for the additional information.
Data type: The PCF structure data type for the additional information.
Included in PCF group: GroupName.
Value: The value to be displayed.


Examples of recording additional activity information

The following examples illustrate how a user application can record additional information when performing an activity on behalf of a trace-route message. In both examples, the WebSphere MQ display route application is used to generate a trace-route message, and display the activity information returned to it.


Example 1

Additional activity information is recorded by a user application in a format where the parameter identifier is not recognized by the WebSphere MQ display route application.

  1. The WebSphere MQ display route application is used to generate and put a trace-route message into a queue manager network. The necessary options are set to request the following:

    • Activity information is accumulated in the message data of the trace-route message.
    • On arrival at the target queue the trace-route message is discarded, and a trace-route reply message is generated and delivered to a specified reply-to queue.
    • On receipt of the trace-route reply message, the WebSphere MQ display route application displays the accumulated activity information.

    The trace-route message is put into the queue manager network.

  2. As the trace-route message is routed through the queue manager network a user application, that is enabled for trace-route messaging, performs a low detail activity on behalf of the message. In addition to writing the standard activity information to the trace-route message, the user application writes the following PCF parameter to the end of the Activity group:

    ColorValue

    Identifier

    65536

    Data type

    MQCFST

    Value

    'Red'
    This additional PCF parameter gives further information about the activity that was performed, however it is written in a format where the parameter identifier is not recognized by the WebSphere MQ display route application.
  3. The trace-route messages reaches the target queue and a trace-route reply message is returned to the WebSphere MQ display route application. The additional activity information is displayed as follows:

     65536: 'Red'
    
    The WebSphere MQ display route application does not recognize the parameter identifier of the PCF parameter and displays it as a numeric value. The context of the additional information is not clear.

    For an example of when the WebSphere MQ display route application does recognize the parameter identifier of the PCF parameter, see Example 2 .


Concept Feedback


Last updated: Monday, September 23, 2013



Example 2

Additional activity information is recorded by a user application in a format where the parameter identifier is recognized by the WebSphere MQ display route application.

  1. TheWebSphere MQ display route application is used to generate and put a trace-route message into a queue manager network in the same fashion as in Example 1 .
  2. As the trace-route message is routed through the queue manager network a user application, that is enabled for trace-route messaging, performs a low detail activity on behalf of the message. In addition to writing the standard activity information to the trace-route message, the user application writes the following PCF parameters to the end of the Activity group:

    ColorInfo

    Color information

    Description Grouped parameters specifying information about a color.
    Identifier: MQGACF_VALUE_NAMING.
    Data type: MQCFGR.
    Parameters in group:
    • ColorName
    • ColorValue

    ColorName

    Color name

    Description Contains the name to be displayed by the WebSphere MQ display route application which puts the value of ColorValue into context.
    Identifier: MQCA_VALUE_NAME.
    Data type: MQCFST.
    Included in PCF group: ColorInfo.
    Value: 'Color'

    ColorValue

    Description Contains the value to be displayed by the WebSphere MQ display route application.
    Identifier: 65536.
    Data type: MQCFST.
    Included in PCF group: ColorInfo.
    Value: 'Red'

    These additional PCF parameters gives further information about the activity that was performed. These PCF parameters are written in a format where the parameter identifier is recognized by the WebSphere MQ display route application.
  3. The trace-route messages reaches the target queue and a trace-route reply message is returned to the WebSphere MQ display route application. The additional activity information is displayed as follows:

     Color: 'Red'
    
    The WebSphere MQ display route application recognizes that the parameter identifier of the PCF structure containing the value of the additional activity information has a corresponding name. The corresponding name is displayed instead of the numeric value.


Concept Feedback


Last updated: Monday, September 23, 2013



WebSphere MQ display route application

Use the WebSphere MQ display route application (dspmqrte) to work with trace-route messages and activity information related to a trace-route message, using a command-line interface.

Note: To run a Client Application against a queue manager, the Client Attachment feature must be installed.

You can use the WebSphere MQ display route application for the following purposes:


Parameters for trace-route messages

Use this page to obtain an overview of the parameters provided by the WebSphere MQ display route application, dspmqrte, to determine the characteristics of a trace-route message, including how it is treated as it is routed through a queue manager network.


Queue manager connection

Use this page to specify the queue manager that the WebSphere MQ display route application connects to

-c
Specifies that the WebSphere MQ display route application connects as a client application.

If you do not specify this parameter, the WebSphere MQ display route application does not connect as a client application.

-m QMgrName
The name of the queue manager to which the WebSphere MQ display route application connects. The name can contain up to 48 characters.

If you do not specify this parameter, the default queue manager is used.


Reference Feedback


Last updated: Monday, September 23, 2013



The target destination

Use this page to specify the target destination of a trace-route message

-q TargetQName
If the WebSphere MQ display route application is being used to send a trace-route message into a queue manager network, TargetQName specifies the name of the target queue.
-ts TargetTopicString
Specifies the topic string.
-qm TargetQMgr
Qualifies the target destination; normal queue manager name resolution will then apply. The target destination is specified with -q TargetQName or -ts TargetTopicString.

If you do not specify this parameter, the queue manager to which the WebSphere MQ display route application is connected is used as the target queue manager.

-o
Specifies that the target destination is not bound to a specific destination. Typically this parameter is used when the trace-route message is to be put across a cluster. The target destination is opened with option MQOO_BIND_NOT_FIXED.

If you do not specify this parameter, the target destination is bound to a specific destination.


Reference Feedback


Last updated: Monday, September 23, 2013



The publication topic

For publish/subscribe applications, use this page to specify the topic string of a trace-route message for the WebSphere MQ display route application to publish

-ts TopicName
Specifies a topic string to which the WebSphere MQ display route application is to publish a trace-route message, and puts this application into topic mode. In this mode, the application traces all of the messages that result from the publish request.

You can also use the WebSphere MQ display route application to display the results from an activity report that was generated for publish messages.


Reference Feedback


Last updated: Monday, September 23, 2013



Message mimicking

Use this page to configure a trace-route message to mimic a message, for example when the original message did not reach its intended destination

One use of trace-route messaging is to help determine the last known location of a message that did not reach its intended destination. The WebSphere MQ display route application provides parameters that can help configure a trace-route message to mimic the original message. When mimicking a message, you can use the following parameters:

-l Persistence
Specifies the persistence of the generated trace-route message. Possible values for Persistence are:

yes

The generated trace-route message is persistent. (MQPER_PERSISTENT).

no

The generated trace-route message is not persistent. (MQPER_NOT_PERSISTENT).

q

The generated trace-route message inherits its persistence value from the destination specified by -q TargetQName or -ts TargetTopicString. (MQPER_PERSISTENCE_AS_Q_DEF).
A trace-route reply message, or any report messages, returned will share the same persistence value as the original trace-route message.

If Persistence is specified as yes , you must specify the parameter -rq ReplyToQ. The reply-to queue must not resolve to a temporary dynamic queue.

If you do not specify this parameter, the generated trace-route message is not persistent.

-p Priority
Specifies the priority of the trace-route message. The value of Priority is either greater than or equal to 0, or MQPRI_PRIORITY_AS_Q_DEF. MQPRI_PRIORITY_AS_Q_DEF specifies that the priority value is taken from the destination specified by -q TargetQName or -ts TargetTopicString.

If you do not specify this parameter, the priority value is taken from the destination specified by -q TargetQName or -ts TargetTopicString.

-xs Expiry
Specifies the expiry time for the trace-route message, in seconds.

If you do not specify this parameter, the expiry time is specified as 60 seconds.

-ro none | ReportOption

none

Specifies no report options are set.

ReportOption

Specifies report options for the trace-route message. Multiple report options can be specified using a comma as a separator. Possible values for ReportOption are:

activity

The report option MQRO_ACTIVITY is set.

coa

The report option MQRO_COA_WITH_FULL_DATA is set.

cod

The report option MQRO_COD_WITH_FULL_DATA is set.

exception

The report option MQRO_EXCEPTION_WITH_FULL_DATA is set.

expiration

The report option MQRO_EXPIRATION_WITH_FULL_DATA is set.

discard

The report option MQRO_DISCARD_MSG is set.

If neither -ro ReportOption nor -ro none are specified, then the MQRO_ACTIVITY and MQRO_DISCARD_MSG report options are specified.

The WebSphere MQ display route application does not allow you to add user data to the trace-route message. If you require user data to be added to the trace-route message you must generate the trace-route message manually.


Reference Feedback


Last updated: Monday, September 23, 2013



Recorded activity information

Use this page to specify the method used to return recorded activity information, which you can then use to determine the route that a trace-route message has taken

Recorded activity information can be returned as follows:

When using dspmqrte, the method used to return recorded activity information is determined using the following parameters:

The activity report option, specified using -ro
Specifies that activity information is returned using activity reports. By default activity recording is enabled.
-ac -ar
Specifies that activity information is accumulated in the trace-route message, and that a trace-route reply message is to be generated.

-ac
Specifies that activity information is to be accumulated within the trace-route message.

If you do not specify this parameter, activity information is not accumulated within the trace-route message.

-ar
Requests that a trace-route reply message containing all accumulated activity information is generated in the following circumstances:

  • The trace-route message is discarded by a WebSphere MQ queue manager.
  • The trace-route message is put to a local queue (target queue or dead-letter queue) by a WebSphere MQ queue manager.
  • The number of activities performed on the trace-route message exceeds the value of specified in -s Activities.

-ac -d yes
Specifies that activity information is accumulated in the trace-route message, and that on arrival, the trace-route message will be put on the target queue.

-ac
Specifies that activity information is to be accumulated within the trace-route message.

If you do not specify this parameter, activity information is not accumulated within the trace-route message.

-d yes
On arrival, the trace-route message is put to the target queue, even if the queue manager does not support trace-route messaging.

If you do not specify this parameter, the trace-route message is not put to the target queue.

The trace-route message can then be retrieved from the target queue, and the recorded activity information acquired.

You can combine these methods as required.

Additionally, the detail level of the recorded activity information can be specified using the following parameter:

-t Detail
Specifies the activities that are recorded. The possible values for Detail are:

low

Activities performed by user-defined application are recorded only.

medium

high

If you do not specify this parameter, medium level activities are recorded.

By default the WebSphere MQ display route application uses a temporary dynamic queue to store the returned messages. When the WebSphere MQ display route application ends, the temporary dynamic queue is closed, and any messages are purged. If the returned messages are required beyond the current execution of the WebSphere MQ display route application ends, then a permanent queue must be specified using the following parameters:

-rq ReplyToQ
Specifies the name of the reply-to queue that all responses to the trace-route message are sent to. If the trace-route message is persistent, or if the -n parameter is specified, a reply-to queue must be specified that is not a temporary dynamic queue.

If you do not specify this parameter then a dynamic reply-to queue is created using the system default model queue, SYSTEM.DEFAULT.MODEL.QUEUE.

-rqm ReplyToQMgr
Specifies the name of the queue manager where the reply-to queue resides. The name can contain up to 48 characters.

If you do not specify this parameter, the queue manager to which the WebSphere MQ display route application is connected is used as the reply-to queue manager.


Reference Feedback


Last updated: Monday, September 23, 2013



How the trace-route message is handled

Use this page to control how a trace-route message is handled as it is routed through a queue manager network.

The following parameters can restrict where the trace-route message can be routed in the queue manager network:

-d Deliver
Specifies whether the trace-route message is to be delivered to the target queue on arrival. Possible values for Deliver are:

If you do not specify this parameter, the trace-route message is not put to the target queue.

-f Forward
Specifies the type of queue manager that the trace-route message can be forwarded to. For details of the algorithm that queue managers use to determine whether to forward a message to a remote queue manager, refer to The TraceRoute PCF group . The possible values for Forward are:

all

The trace-route message is forwarded to any queue manager.

Warning: If forwarded to a WebSphere MQ queue manager earlier than Version 6.0, the trace-route message will not be recognized and can be delivered to a local queue despite the value of the -d Deliver parameter.

supported

The trace-route message is only forwarded to a queue manager that will honor the Deliver parameter from the TraceRoute PCF group

If you do not specify this parameter, the trace-route message will only be forwarded to a queue manager that will honor the Deliver parameter.

The following parameters can prevent a trace-route message from remaining in a queue manager network indefinitely:

-s Activities
Specifies the maximum number of recorded activities that can be performed on behalf of the trace-route message before it is discarded. This prevents the trace-route message from being forwarded indefinitely if caught in an infinite loop. The value of Activities is either greater than or equal to 1, or MQROUTE_UNLIMITED_ACTIVITIES. MQROUTE_UNLIMITED_ACTIVITIES specifies that an unlimited number of activities can be performed on behalf of the trace-route message.

If you do not specify this parameter, an unlimited number of activities can be performed on behalf of the trace-route message.

-xs Expiry
Specifies the expiry time for the trace-route message, in seconds.

If you do not specify this parameter, the expiry time is specified as 60 seconds.

-xp PassExpiry
Specifies whether the expiry time from the trace-route message is passed on to a trace-route reply message. Possible values for PassExpiry are:

yes

The report option MQRO_PASS_DISCARD_AND_EXPIRY is specified in the message descriptor of the trace-route message.

If a trace-route reply message, or activity reports, are generated for the trace-route message, the MQRO_DISCARD report option (if specified), and the remaining expiry time are passed on.

This is the default value.

no

The report option MQRO_PASS_DISCARD_AND_EXPIRY is not specified.

If a trace-route reply message is generated for the trace-route message, the discard option and expiry time from the trace-route message are not passed on.

If you do not specify this parameter, MQRO_PASS_DISCARD_AND_EXPIRY is not specified.
The discard report option, specified using -ro
Specifies the MQRO_DISCARD_MSG report option. This can prevent the trace-route message remaining in the queue manager network indefinitely.


Reference Feedback


Last updated: Monday, September 23, 2013



Display of activity information

The WebSphere MQ display route application can display activity information for a trace-route message that it has just put into a queue manager network, or it can display activity information for a previously generated trace-route message. It can also display additional information recorded by user-written applications.

To specify whether activity information returned for a trace-route message is displayed, specify the following parameter:

-n
Specifies that activity information returned for the trace-route message is not to be displayed.

If this parameter is accompanied by a request for a trace-route reply message, (-ar), or any of the report generating options from (-ro ReportOption), then a specific (non-model) reply-to queue must be specified using -rq ReplyToQ. By default, only activity report messages are requested.

After the trace-route message is put to the specified target queue, a 48 character hexadecimal string is displayed containing the message identifier of the trace-route message. The message identifier can be used by the WebSphere MQ display route application to display the activity information for the trace-route message at a later time, using the -i CorrelId parameter.

If you do not specify this parameter, activity information returned for the trace-route message is displayed in the form specified by the -v parameter.

When displaying activity information for a trace-route message that has just been put into a queue manager network, the following parameter can be specified:

-w WaitTime
Specifies the time, in seconds, that the WebSphere MQ display route application will wait for activity reports, or a trace-route reply message, to return to the specified reply-to queue.

If you do not specify this parameter, the wait time is specified as the expiry time of the trace-route message, plus 60 seconds.

When displaying previously accumulated activity information the following parameters must be set:

-q TargetQName
If the WebSphere MQ display route application is being used to view previously gathered activity information, TargetQName specifies the name of the queue where the activity information is stored.
-i CorrelId
This parameter is used when the WebSphere MQ display route application is used to display previously accumulated activity information only. There can be many activity reports and trace-route reply messages on the queue specified by -q TargetQName. CorrelId is used to identify the activity reports, or a trace-route reply message, related to a trace-route message. Specify the message identifier of the original trace-route message in CorrelId.

The format of CorrelId is a 48 character hexadecimal string.

The following parameters can be used when displaying previously accumulated activity information, or when displaying current activity information for a trace-route message:

-b
Specifies that the WebSphere MQ display route application will only browse activity reports or a trace-route reply message related to a message. This allows activity information to be displayed again at a later time.

If you do not specify this parameter, the WebSphere MQ display route application will destructively get activity reports or a trace-route reply message related to a message.

-v summary | all | none | outline DisplayOption

summary

The queues that the trace-route message was routed through are displayed.

all

All available information is displayed.

none

No information is displayed.

outline DisplayOption

Specifies display options for the trace-route message. Multiple display options can be specified using a comma as a separator.

If no values are supplied the following is displayed:

  • The application name
  • The type of each operation
  • Any operation specific parameters

Possible values for DisplayOption are:

activity

All non-PCF group parameters in Activity PCF groups are displayed.

identifiers

Values with parameter identifiers MQBACF_MSG_ID or MQBACF_CORREL_ID are displayed. This overrides msgdelta.

message

All non-PCF group parameters in Message PCF groups are displayed. When this value is specified, you cannot specify msgdelta.

msgdelta

All non-PCF group parameters in Message PCF groups, that have changed since the last operation, are displayed. When this value is specified, you cannot specify message.

operation

All non-PCF group parameters in Operation PCF groups are displayed.

traceroute

All non-PCF group parameters in TraceRoute PCF groups are displayed.

If you do not specify this parameter, a summary of the message route is displayed.


Display of additional information

As a trace-route message is routed through a queue manager network, user-written applications can record additional information by writing one or more additional PCF parameters to the message data of the trace-route message or to the message data of an activity report. For the WebSphere MQ display route application to display additional information in a readable form it must be recorded in a specific format, as described in Additional activity information .


Reference Feedback


Last updated: Monday, September 23, 2013



WebSphere MQ display route application examples

The following examples show how you can use the WebSphere MQ display route application. In each example, two queue managers (QM1 and QM2) are inter-connected by two channels (QM2.TO.QM1 and QM1.TO.QM2).


Example 1 - Requesting activity reports

Display activity information from a trace-route message delivered to the target queue

In this example the WebSphere MQ display route application connects to queue manager, QM1, and is used to generate and deliver a trace-route message to the target queue, TARGET.Q, on remote queue manager, QM2. The necessary report option is specified so that activity reports are requested as the trace-route reply message is routed. On arrival at the target queue the trace-route message is discarded. Activity information returned to the WebSphere MQ display route application using activity reports is put in order and displayed.

Figure 1. Requesting activity reports, Diagram 1

Figure 2. Requesting activity reports, Diagram 2

Figure 3. Requesting activity reports, Diagram 3

Figure 4. Requesting activity reports, Diagram 4


Concept Feedback


Last updated: Monday, September 23, 2013



Example 2 - Requesting a trace-route reply message

Generate and deliver a trace-route message to the target queue

In this example the WebSphere MQ display route application connects to queue manager, QM1, and is used to generate and deliver a trace-route message to the target queue, TARGET.Q, on remote queue manager, QM2. The necessary option is specified so that activity information is accumulated in the trace-route message. On arrival at the target queue a trace-route reply message is requested, and the trace-route message is discarded.

Figure 1. Requesting a trace-route reply message, Diagram 1

Figure 2. Requesting a trace-route reply message, Diagram 2

Figure 3. Requesting a trace-route reply message, Diagram 3

Figure 4. Requesting a trace-route reply message, Diagram 4

The output that is displayed follows:

 AMQ8653: DSPMQRTE command started with options '-m QM1 -q TARG.AT.QM2 -rq
 TR.REPLY.Q'.
AMQ8659: DSPMQRTE command successfully put a message on queue 'QM2', queue  manager 'QM1'.
AMQ8674: DSPMQRTE command is now waiting for information to display.
AMQ8666: Queue 'QM2' on queue manager 'QM1'.
AMQ8666: Queue 'TARGET.Q' on queue manager 'QM2'.
AMQ8652: DSPMQRTE command has finished.  


Concept Feedback


Last updated: Monday, September 23, 2013



Example 3 - Delivering activity reports to the system queue

Detect when activity reports are delivered to queues other than the reply-to queue and use the WebSphere MQ display route application to read activity reports from the other queue.

This example is the same as Example 1 - Requesting activity reports , except that QM2 now has the value of the ACTIVREC queue manage attribute set to QUEUE. Channel QM1.TO.QM2 must have been restarted for this to take effect.

This example demonstrates how to detect when activity reports are delivered to queues other than the reply-to queue. Once detected, the WebSphere MQ display route application is used to read activity reports from another queue.

Figure 1. Delivering activity reports to the system queue, Diagram 1

The output that is displayed follows:

 AMQ8653: DSPMQRTE command started with options '-m QM1 -q TARG.AT.QM2 -rq
          ACTIV.REPLY.Q -v outline identifiers'.
AMQ8659: DSPMQRTE command successfully put a message on queue 'QM2', queue           manager 'QM1'.
AMQ8674: DSPMQRTE command is now waiting for information to display.  
--------------------------------------------------------------------------------
Activity:
 ApplName: 'cann\output\bin\dspmqrte.exe'
 
 Operation:
  OperationType: Put

  Message:

   MQMD:
    MsgId: X'414D51204C4152474551202020202020A3C9154220001502'
    CorrelId: X'414D51204C4152474551202020202020A3C9154220001503'
  QMgrName: 'QM1                                             '
  QName: 'TARG.AT.QM2                                     '
  ResolvedQName: 'QM2                                             '
  RemoteQName: 'TARGET.Q                                        '
  RemoteQMgrName: 'QM2                                             ' 
--------------------------------------------------------------------------------
Activity:
 ApplName: 'cann\output\bin\runmqchl.EXE'

 Operation:
  OperationType: Get

  Message:

   MQMD:
    MsgId: X'414D51204C4152474551202020202020A3C9154220001505'
    CorrelId: X'414D51204C4152474551202020202020A3C9154220001502'

   EmbeddedMQMD:
    MsgId: X'414D51204C4152474551202020202020A3C9154220001502'
    CorrelId: X'414D51204C4152474551202020202020A3C9154220001503'
  QMgrName: 'QM1                                             '
  QName: 'QM2                                             '
  ResolvedQName: 'QM2                                             '

 Operation:
  OperationType: Send

  Message:

   MQMD:
    MsgId: X'414D51204C4152474551202020202020A3C9154220001502'
    CorrelId: X'414D51204C4152474551202020202020A3C9154220001503'
  QMgrName: 'QM1                                             '
  RemoteQMgrName: 'QM2                                             '
  ChannelName: 'QM1.TO.QM2          '
  ChannelType: Sender
  XmitQName: 'QM2                                             '  
--------------------------------------------------------------------------------
AMQ8652: DSPMQRTE command has finished.  

The output that is displayed follows:

 AMQ8653: DSPMQRTE command started with options '-m QM2 
         -q SYSTEM.ADMIN.ACTIVITY.QUEUE 
         -i 414D51204C4152474551202020202020A3C915420001502 -v outline'.
AMQ8674: DSPMQRTE command is now waiting for information to display.
--------------------------------------------------------------------------------

Activity:
 Activity information unavailable.

--------------------------------------------------------------------------------
Activity:
 ApplName: 'cann\output\bin\AMQRMPPA.EXE'

 Operation:
  OperationType: Receive
  QMgrName: 'QM2                                             '
  RemoteQMgrName: 'QM1                                             '
  ChannelName: 'QM1.TO.QM2          '
  ChannelType: Receiver

 Operation:
  OperationType: Discard
  QMgrName: 'QM2                                             '
  QName: 'TARGET.Q                                        '
  Feedback: NotDelivered

--------------------------------------------------------------------------------
AMQ8652: DSPMQRTE command has finished.


Concept Feedback


Last updated: Monday, September 23, 2013



Example 4 - Diagnosing a channel problem

Diagnose a problem in which the trace-route message does not reach the target queue

In this example the WebSphere MQ display route application connects to queue manager, QM1, generates a trace-route message, then attempts to deliver it to the target queue, TARGET.Q, on remote queue manager, QM2. In this example the trace-route message does not reach the target queue. The available activity report is used to diagnose the problem.

Figure 1. Diagnosing a channel problem

The output that is displayed follows:

 AMQ8653: DSPMQRTE command started with options '-m QM1 -q TARG.AT.QM2
          -rq ACTIV.REPLY.Q -v outline'.
AMQ8659: DSPMQRTE command successfully put a message on queue 'QM2',
          queue manager 'QM1'.
AMQ8674: DSPMQRTE command is now waiting for information to display.
--------------------------------------------------------------------------------
Activity:  
 ApplName: 'cann\output\bin\dspmqrte.exe'

 Operation:
  OperationType: Put
  QMgrName: 'QM1                                             '
  QName: 'TARG.AT.QM2                                     '
  ResolvedQName: 'QM2                                             '
  RemoteQName: 'TARGET.Q                                        '
  RemoteQMgrName: 'QM2                                             '

--------------------------------------------------------------------------------
 AMQ8652: DSPMQRTE command has finished. 


Concept Feedback


Last updated: Monday, September 23, 2013



Activity report reference

Use this page to obtain an overview of the activity report message format. The activity report message data contains the parameters that describe the activity.


Activity report format

Activity reports are standard WebSphere MQ report messages containing a message descriptor and message data. Activity reports are PCF messages generated by applications that have performed an activity on behalf of a message as it has been routed through a queue manager network.

Activity reports contain the following information:

A message descriptor

An MQMD structure

Message data

Consists of the following:

  • An embedded PCF header (MQEPH).
  • Activity report message data.

Activity report message data consists of the Activity PCF group and, if generated for a trace-route message, the TraceRoute PCF group.

Table 1 shows the structure of these reports, including parameters that are returned only under certain conditions.

Activity report format

MQMD structure Embedded PCF header MQEPH structure Activity report message data

Structure identifier
Structure version
Report options
Message type
Expiration time
Feedback
Encoding
Coded character set ID
Message format
Priority
Persistence
Message identifier
Correlation identifier
Backout count
Reply-to queue
Reply-to queue manager
User identifier
Accounting token
Application identity data
Application type
Application name
Put date
Put time
Application origin data
Group identifier
Message sequence number
Offset
Message flags
Original length
 

Structure identifier
Structure version
Structure length
Encoding
Coded character set ID
Message format
Flags
PCF header (MQCFH)
  Structure type
  Structure length
  Structure version
  Command identifier
  Message sequence number
  Control options
  Completion code
  Reason code
  Parameter count
 

Activity
  Activity application name
  Activity application type
  Activity description
  Operation
    Operation type
    Operation date
    Operation time
    Message
      Message length
      MQMD 8
      EmbeddedMQMD
    Queue manager name
    Queue sharing group name
    Queue name 1 2 3 7
    Resolved queue name 1 3 7
    Remote queue name 3 7
    Remote queue manager name 2 3 4 5 7
    Subscription level 9
    Subscription identifier 9
    Feedback 2 10     
    Channel name 4 5
    Channel type 4 5
    Transmission queue name 5
  TraceRoute 6
    Detail
    Recorded activities
    Unrecorded activities
    Discontinuity count
    Max activities
    Accumulate
    Deliver
 

Notes:

  1. Returned for Get and Browse operations.
  2. Returned for Discard operations.
  3. Returned for Put, Put Reply, and Put Report operations.
  4. Returned for Receive operations.
  5. Returned for Send operations.
  6. Returned for trace-route messages.
  7. Not returned for Put operations to a topic, contained within Publish activities.
  8. Not returned for Excluded Publish operations. For Publish and Discarded Publish operations, returned containing a subset of parameters.
  9. Returned for Publish, Discarded Publish, and Excluded Publish operations.
  10. Returned for Discarded Publish and Excluded Publish operations.


Concept Feedback


Last updated: Monday, September 23, 2013



Activity report MQMD (message descriptor)

Use this page to view the values contained by the MQMD structure for an activity report

StrucId

Structure identifier:

Data type

MQCHAR4

Value

MQMD_STRUC_ID.

Version

Structure version number

Data type

MQLONG

Values

Copied from the original message descriptor. Possible values are:

MQMD_VERSION_1

Version-1 message descriptor structure, supported in all environments.

MQMD_VERSION_2

Version-2 message descriptor structure, supported on AIX, HP-UX, z/OS, HP OpenVMS, IBM i, Solaris, Linux, Windows, and all WebSphere MQ MQI clients connected to these systems.

Report

Options for further report messages

Data type

MQLONG

Value

If MQRO_PASS_DISCARD_AND_EXPIRY or MQRO_DISCARD_MSG were specified in the Report field of the original message descriptor:

MQRO_DISCARD

The report is discarded if it cannot be delivered to the destination queue.
Otherwise:

MQRO_NONE

No reports required.

MsgType

Indicates type of message

Data type

MQLONG

Value

MQMT_REPORT

Expiry

Report message lifetime

Data type

MQLONG

Value

If the Report field in the original message descriptor is specified as MQRO_PASS_DISCARD_AND_EXPIRY, the remaining expiry time from the original message is used.

Otherwise:

MQEI_UNLIMITED

The report does not have an expiry time.

Feedback

Encoding

CodedCharSetId

Format

Priority

Persistence

MsgId

CorrelId

BackoutCount

ReplyToQ

ReplyToQMgr

UserIdentifier

AccountingToken

ApplIdentityData

PutApplType

PutApplName

PutDate

PutTime

ApplOriginData

If Version is MQMD_VERSION_2, the following additional fields are present:

GroupId

MsgSeqNumber

Offset

MsgFlags

OriginalLength


Reference Feedback


Last updated: Monday, September 23, 2013



Activity report MQEPH (Embedded PCF header)

Use this page to view the values contained by the MQEPH structure for an activity report

The MQEPH structure contains a description of both the PCF information that accompanies the message data of an activity report, and the application message data that follows it.

For an activity report, the MQEPH structure contains the following values:

StrucId

Version

StrucLength

Encoding

CodedCharSetId

Format

Flags

PCFHeader


Reference Feedback


Last updated: Monday, September 23, 2013



Activity report MQCFH (PCF header)

Use this page to view the PCF values contained by the MQCFH structure for an activity report

For an activity report, the MQCFH structure contains the following values:

Type

StrucLength

Version

Command

MsgSeqNumber

Control

CompCode

Reason

ParameterCount


Reference Feedback


Last updated: Monday, September 23, 2013



Activity report message data

Use this page to view the parameters contained by the Activity PCF group in an activity report message. Some parameters are returned only when specific operations have been performed.

Activity report message data consists of the Activity PCF group and, if generated for a trace-route message, the TraceRoute PCF group. The Activity PCF group is detailed in this topic.

Some parameters, which are described as Operation-specific activity report message data , are returned only when specific operations have been performed.

For an activity report, the activity report message data contains the following parameters:

Activity

ActivityApplName

ActivityApplType

ActivityDescription

Operation

OperationType

OperationDate

OperationTime

Message

MsgLength

MQMD

EmbeddedMQMD

StrucId

Version

Report

MsgType

Expiry

Feedback

Encoding

CodedCharSetId

Format

Priority

Persistence

MsgId

CorrelId

BackoutCount

ReplyToQ

ReplyToQMgr

UserIdentifier

AccountingToken

ApplIdentityData

PutApplType

PutApplName

PutDate

PutTime

ApplOriginData

GroupId

MsgSeqNumber

Offset

MsgFlags

OriginalLength

QMgrName

QSGName

TraceRoute

The values of the parameters in the TraceRoute PCF group are those from the trace-route message at the time the activity report was generated.


Reference Feedback


Last updated: Monday, September 23, 2013



Operation-specific activity report message data

Use this page to view the additional PCF parameters that might be returned in the PCF group Operation in an activity report, depending on the value of the OperationType parameter

The additional parameters vary depending on the following operation types:


Get/Browse (MQOPER_GET/MQOPER_BROWSE)

The additional activity report message data parameters that are returned in the PCF group Operation for the Get/Browse (MQOPER_GET/MQOPER_BROWSE) operation type (a message on a queue was got, or browsed).

QName

ResolvedQName


Reference Feedback


Last updated: Monday, September 23, 2013



Discard (MQOPER_DISCARD)

The additional activity report message data parameters that are returned in the PCF group Operation for the Discard (MQOPER_DISCARD) operation type (a message was discarded).

Feedback

QName

RemoteQMgrName


Reference Feedback


Last updated: Monday, September 23, 2013



Publish/Discarded Publish/Excluded Publish (MQOPER_PUBLISH/MQOPER_DISCARDED_PUBLISH/MQOPER_EXCLUDED_PUBLISH)

The additional activity report message data parameters that are returned in the PCF group Operation for the Publish/Discarded Publish/Excluded Publish (MQOPER_PUBLISH/MQOPER_DISCARDED_PUBLISH/MQOPER_EXCLUDED_PUBLISH) operation type (a publish/subscribe message was delivered, discarded, or excluded).

SubId

SubLevel

Feedback

The Publish operation MQOPER_PUBLISH provides information about a message delivered to a particular subscriber. This operation describes the elements of the onward message that might have changed from the message described in the associated Put operation. Similarly to a Put operation, it contains a message group MQGACF_MESSAGE and, inside that, an MQMD group MQGACF_MQMD. However, this MQMD group contains only the following fields, which can be overridden by a subscriber: Format, Priority, Persistence, MsgId, CorrelId, UserIdentifier, AccountingToken, ApplIdentityData.

The SubId and SubLevel of the subscriber are included in the operation information. You can use the SubID with the MQCMD_INQUIRE_SUBSCRIBER PCF command to retrieve all other attributes for a subscriber.

The Discarded Publish operation MQOPER_DISCARDED_PUBLISH is analogous to the Discard operation that is used when a message is not delivered in point-to-point messaging. A message is not delivered to a subscriber if the message was explicitly requested not to be delivered to a local destination and this subscriber specifies a local destination. A message is also considered not delivered if there is a problem getting the message to the destination queue, for example, because the queue is full.

The information in a Discarded Publish operation is the same as for a Publish operation, with the addition of a Feedback field that gives the reasons why the message was not delivered. This feedback field contains MQFB_* or MQRC_* values that are common with the MQOPER_DISCARD operation. The reason for discarding a publish, as opposed to excluding it, are the same as the reasons for discarding a put.

The Excluded Publish operation MQOPER_EXCLUDED_PUBLISH provides information about a subscriber that was considered for delivery of the message, because the topic on which the subscriber is subscribing matches that of the associated Put operation, but the message was not delivered to the subscriber because other selection criteria do not match with the message that is being put to the topic. As with a Discarded Publish operation, the Feedback field provides information about the reason why this subscription was excluded. However, unlike the Discarded Publish operation, no message-related information is provided because no message was generated for this subscriber.


Reference Feedback


Last updated: Monday, September 23, 2013



Put/Put Reply/Put Report (MQOPER_PUT/MQOPER_PUT_REPLY/MQOPER_PUT_REPORT)

The additional activity report message data parameters that are returned in the PCF group Operation for the Put/Put Reply/Put Report (MQOPER_PUT/MQOPER_PUT_REPLY/MQOPER_PUT_REPORT) operation type (a message, reply message, or report message was put to a queue).

QName

ResolvedQName

RemoteQName

RemoteQMgrName

TopicString

Feedback


Reference Feedback


Last updated: Monday, September 23, 2013



Receive (MQOPER_RECEIVE)

The additional activity report message data parameters that are returned in the PCF group Operation for the Receive (MQOPER_RECEIVE) operation type (a message was received on a channel).

ChannelName

ChannelType

RemoteQMgrName


Reference Feedback


Last updated: Monday, September 23, 2013



Send (MQOPER_SEND)

The additional activity report message data parameters that are returned in the PCF group Operation for the Send (MQOPER_SEND) operation type (a message was sent on a channel).

ChannelName

ChannelType

XmitQName

RemoteQMgrName


Reference Feedback


Last updated: Monday, September 23, 2013



Trace-route message reference

Use this page to obtain an overview of the trace-route message format. The trace-route message data includes parameters that describe the activities that the trace-route message has caused


Trace-route message format

Trace-route messages are standard WebSphere MQ messages containing a message descriptor and message data. The message data contains information about the activities performed on a trace-route message as it has been routed through a queue manager network.

Trace-route messages contain the following information:

A message descriptor

An MQMD structure, with the Format field set to MQFMT_ADMIN or MQFMT_EMBEDDED_PCF.

Message data

Consists of either:

  • A PCF header (MQCFH) and trace-route message data, if Format is set to MQFMT_ADMIN, or
  • An embedded PCF header (MQEPH), trace-route message data, and additional user-specified message data, if Format is set to MQFMT_EMBEDDED_PCF.

When using the WebSphere MQ display route application to generate a trace-route message, Format is set to MQFMT_ADMIN.

The content of the trace-route message data is determined by the Accumulate parameter from the TraceRoute PCF group, as follows:

Table 1 shows the structure of a trace-route message.

Trace-route message format

MQMD structure Embedded PCF header MQEPH structure Trace-route message data

Structure identifier
Structure version
Report options
Message type
Expiration time
Feedback
Encoding
Coded character set ID
Message format
Priority
Persistence
Message identifier
Correlation identifier
Backout count
Reply-to queue
Reply-to queue manager
User identifier
Accounting token
Application identity data
Application type
Application name
Put date
Put time
Application origin data
Group identifier
Message sequence number
Offset
Message flags
Original length
 

Structure identifier
Structure version
Structure length
Encoding
Coded character set ID
Message format
Flags
PCF header (MQCFH)
  Structure type
  Structure length
  Structure version
  Command identifier
  Message sequence number
  Control options
  Completion code
  Reason code
  Parameter count
 

TraceRoute
  Detail
  Recorded activities
  Unrecorded activities
  Discontinuity count
  Max activities
  Accumulate
  Deliver
 


Concept Feedback


Last updated: Monday, September 23, 2013



Trace-route message MQMD (message descriptor)

Use this page to view the values contained by the MQMD structure for a trace-route message

StrucId

Version

Report

MsgType

Expiry

Feedback

Encoding

CodedCharSetId

Format

Priority

Persistence

MsgId

CorrelId

BackoutCount

ReplyToQ

ReplyToQMgr

UserIdentifier

AccountingToken

ApplIdentityData

PutApplType

PutApplName

PutDate

PutTime

ApplOriginData


Reference Feedback


Last updated: Monday, September 23, 2013



Trace-route message MQEPH (Embedded PCF header)

Use this page to view the values contained by the MQEPH structure for a trace-route message

The MQEPH structure contains a description of both the PCF information that accompanies the message data of a trace-route message, and the application message data that follows it. An MQEPH structure is used only if additional user message data follows the TraceRoute PCF group.

For a trace-route message, the MQEPH structure contains the following values:

StrucId

Version

StrucLength

Encoding

CodedCharSetId

Format

Flags

PCFHeader


Reference Feedback


Last updated: Monday, September 23, 2013



Trace-route message MQCFH (PCF header)

Use this page to view the PCF values contained by the MQCFH structure for a trace-route message

For a trace-route message, the MQCFH structure contains the following values:

Type

StrucLength

Version

Command

MsgSeqNumber

Control

CompCode

Reason

ParameterCount


Reference Feedback


Last updated: Monday, September 23, 2013



Trace-route message data

Use this page to view the parameters that make up the TraceRoute PCF group part of trace-route message data

The content of trace-route message data depends on the Accumulate parameter from the TraceRoute PCF group. Trace-route message data consists of the TraceRoute PCF group, and zero or more Activity PCF groups. The TraceRoute PCF group is detailed in this topic. Refer to the related information for details of the Activity PCF group.

Trace-route message data contains the following parameters:

TraceRoute

Detail

RecordedActivities

UnrecordedActivities

DiscontinuityCount

MaxActivities

Accumulate

Forward

Deliver


Reference Feedback


Last updated: Monday, September 23, 2013



Trace-route reply message reference

Use this page to obtain an overview of the trace-route reply message format. The trace-route reply message data is a duplicate of the trace-route message data from the trace-route message for which it was generated


Trace-route reply message format

Trace-route reply messages are standard WebSphere MQ messages containing a message descriptor and message data. The message data contains information about the activities performed on a trace-route message as it has been routed through a queue manager network.

Trace-route reply messages contain the following information:

A message descriptor

An MQMD structure

Message data

A PCF header (MQCFH) and trace-route reply message data
Trace-route reply message data consists of one or more Activity PCF groups.

When a trace-route message reaches its target queue, a trace-route reply message can be generated that contains a copy of the activity information from the trace-route message. The trace-route reply message will be delivered to a reply-to queue or to a system queue.

Table 1 shows the structure of a trace-route reply message, including parameters that are only returned under certain conditions.

Trace-route reply message format

MQMD structure PCF header MQCFH structure Trace-route reply message data

Structure identifier
Structure version
Report options
Message type
Expiration time
Feedback
Encoding
Coded character set ID
Message format
Priority
Persistence
Message identifier
Correlation identifier
Backout count
Reply-to queue
Reply-to queue manager
User identifier
Accounting token
Application identity data
Application type
Application name
Put date
Put time
Application origin data
Group identifier
Message sequence number
Offset
Message flags
Original length
 

PCF header (MQCFH)
  Structure type
  Structure length
  Structure version
  Command identifier
  Message sequence number
  Control options
  Completion code
  Reason code
  Parameter count
 

Activity
  Activity application name
  Activity application type
  Activity description
  Operation
    Operation type
    Operation date
    Operation time
    Message
      Message length
      MQMD
      EmbeddedMQMD
    Queue manager name
    Queue sharing group name
    Queue name 1 2 3
    Resolved queue name 1 3
    Remote queue name 3
    Remote queue manager-
       name 2 3 4 5
    Feedback 2
    Channel name 4 5
    Channel type 4 5
    Transmission queue name 5
  TraceRoute
    Detail
    Recorded activities
    Unrecorded activities
    Discontinuity count
    Max activities
    Accumulate
    Deliver
 

Note:

  1. Returned for Get and Browse operations.
  2. Returned for Discard operations.
  3. Returned for Put, Put Reply, and Put Report operations.
  4. Returned for Receive operations.
  5. Returned for Send operations.


Concept Feedback


Last updated: Monday, September 23, 2013



Trace-route reply message MQMD (message descriptor)

Use this page to view the values contained by the MQMD structure for a trace-route reply message

For a trace-route reply message, the MQMD structure contains the parameters described in Activity report message descriptor . Some of the parameter values in a trace-route reply message descriptor are different from those in an activity report message descriptor, as follows:

MsgType

Feedback

Encoding

CodedCharSetId

Format


Reference Feedback


Last updated: Monday, September 23, 2013



Trace-route reply message MQCFH (PCF header)

Use this page to view the PCF values contained by the MQCFH structure for a trace-route reply message

The PCF header (MQCFH) for a trace-route reply message is the same as for a trace-route message.


Reference Feedback


Last updated: Monday, September 23, 2013



Trace-route reply message data

The trace-route reply message data is a duplicate of the trace-route message data from the trace-route message for which it was generated

The trace-route reply message data contains one or more Activity groups. The parameters are described in Activity report message data .


Concept Feedback


Last updated: Monday, September 23, 2013



Accounting and statistics messages

Queue managers generate accounting and statistics messages to record information about the MQI operations performed by WebSphere MQ applications, or to record information about the activities occurring in a WebSphere MQ system.

Accounting messages

Accounting messages are used to record information about the MQI operations performed by WebSphere MQ applications, see Accounting messages .

Statistics messages

Statistics messages are used to record information about the activities occurring in a WebSphere MQ system, see Statistics messages .

Accounting and statistics messages are delivered to one of two system queues. User applications can retrieve the messages from these system queues and use the recorded information for various purposes:


Accounting messages

Accounting messages record information about the MQI operations performed by WebSphere MQ applications. An accounting message is a PCF message that contains a number of PCF structures.

When an application disconnects from a queue manager, an accounting message is generated and delivered to the system accounting queue (SYSTEM.ADMIN.ACCOUNTING.QUEUE). For long running WebSphere MQ applications, intermediate accounting messages are generated as follows:

Accounting messages are in the following categories:

MQI accounting messages

MQI accounting messages contain information relating to the number of MQI calls made using a connection to a queue manager.

Queue accounting messages

Queue accounting messages contain information relating to the number of MQI calls made using connections to a queue manager, grouped by queue.

Each queue accounting message can contain up to 100 records, with every record relating to an activity performed by the application with respect to a specific queue.

Accounting messages are recorded only for local queues. If an application makes an MQI call against an alias queue, the accounting data is recorded against the base queue, and, for a remote queue, the accounting data is recorded against the transmission queue.


Accounting message format

Accounting messages comprise a set of PCF fields that consist of a message descriptor and message data.

Message descriptor

  • An accounting message MQMD (message descriptor)

Accounting message data

  • An accounting message MQCFH (PCF header)
  • Accounting message data that is always returned
  • Accounting message data that is returned if available

The accounting message MQCFH (PCF header) contains information about the application, and the interval for which the accounting data was recorded.

Accounting message data comprises PCF parameters that store the accounting information. The content of accounting messages depends on the message category as follows:

MQI accounting message

MQI accounting message data consists of a number of PCF parameters, but no PCF groups.

Queue accounting message

Queue accounting message data consists of a number of PCF parameters, and in the range 1 through 100 QAccountingData PCF groups.

There is one QAccountingData PCF group for every queue that had accounting data collected. If an application accesses more than 100 queues, multiple accounting messages are generated. Each message has the SeqNumber in the MQCFH (PCF header) updated accordingly, and the last message in the sequence has the Control parameter in the MQCFH specified as MQCFC_LAST.


Concept Feedback


Last updated: Monday, September 23, 2013



Accounting information collection

Use queue and queue manager attributes to control the collection of accounting information. You can also use MQCONNX options to control collection at the connection level.


MQI accounting information

Use the queue manager attribute ACCTMQI to control the collection of MQI accounting information

To change the value of this attribute, use the MQSC command, ALTER QMGR, and specify the parameter ACCTMQI. Accounting messages are generated only for connections that begin after accounting is enabled. The ACCTMQI parameter can have the following values:

ON

MQI accounting information is collected for every connection to the queue manager.

OFF

MQI accounting information is not collected. This is the default value.

For example, to enable MQI accounting information collection use the following MQSC command:

 ALTER QMGR ACCTMQI(ON)


Concept Feedback


Last updated: Monday, September 23, 2013



Command event generation

Use this page to view the situations that cause command events to be generated and to understand the circumstances in which command events are not generated

A command event message is generated in the following situations:

If a command runs against the command event queue, SYSTEM.ADMIN.COMMAND.EVENT, a command event is generated if the queue still exists and it is not put-inhibited.


When command events are not generated

A command event message is not generated in the following circumstances:


Concept Feedback


Last updated: Monday, September 23, 2013



MQCONNX options

Use the ConnectOpts parameter on the MQCONNX call to modify the collection of both MQI and queue accounting information at the connection level by overriding the effective values of the queue manager attributes ACCTMQI and ACCTQ

The ConnectOpts parameter can have the following values:

MQCNO_ACCOUNTING_MQI_ENABLED

If the value of the queue manager attribute ACCTMQI is specified as OFF, MQI accounting is enabled for this connection. This is equivalent of the queue manager attribute ACCTMQI being specified as ON.

If the value of the queue manager attribute ACCTMQI is not specified as OFF, this attribute has no effect.

MQCNO_ACCOUNTING_MQI_DISABLED

If the value of the queue manager attribute ACCTMQI is specified as ON, MQI accounting is disabled for this connection. This is equivalent of the queue manager attribute ACCTMQI being specified as OFF.

If the value of the queue manager attribute ACCTMQI is not specified as ON, this attribute has no effect.

MQCNO_ACCOUNTING_Q_ENABLED

If the value of the queue manager attribute ACCTQ is specified as OFF, queue accounting is enabled for this connection. All queues with ACCTQ specified as QMGR, are enabled for queue accounting. This is equivalent of the queue manager attribute ACCTQ being specified as ON.

If the value of the queue manager attribute ACCTQ is not specified as OFF, this attribute has no effect.

MQCNO_ACCOUNTING_Q_DISABLED

If the value of the queue manager attribute ACCTQ is specified as ON, queue accounting is disabled for this connection. This is equivalent of the queue manager attribute ACCTQ being specified as OFF.

If the value of the queue manager attribute ACCTQ is not specified as ON, this attribute has no effect.

These overrides are by disabled by default. To enable them, set the queue manager attribute ACCTCONO to ENABLED. To enable accounting overrides for individual connections use the following MQSC command:

 ALTER QMGR ACCTCONO(ENABLED)


Concept Feedback


Last updated: Monday, September 23, 2013



Accounting message generation

Accounting messages are generated when an application disconnects from the queue manager. Intermediate accounting messages are also written for long running WebSphere MQ applications.

Accounting messages are generated in either of the following ways when an application disconnects:

Intermediate accounting messages are written for long running WebSphere MQ applications when the interval since the connection was established or since the last intermediate accounting message that was written exceeds the configured interval. The queue manager attribute, ACCTINT, specifies the time, in seconds, after which intermediate accounting messages can be automatically written. Accounting messages are generated only when the application interacts with the queue manager, so applications that remain connected to the queue manager for long periods without executing MQI requests do not generate accounting messages until the execution of the first MQI request following the completion of the accounting interval.

The default accounting interval is 1800 seconds (30 minutes). For example, to change the accounting interval to 900 seconds (15 minutes) use the following MQSC command:

 ALTER QMGR ACCTINT(900)


Concept Feedback


Last updated: Monday, September 23, 2013



Statistics messages

Statistics messages record information about the activities occurring in a WebSphere MQ system. An statistics messages is a PCF message that contains a number of PCF structures.

Statistics messages are delivered to the system queue (SYSTEM.ADMIN.STATISTICS.QUEUE) at configured intervals.

Statistics messages are in the following categories:

MQI statistics messages

MQI statistics messages contain information relating to the number of MQI calls made during a configured interval. For example, the information can include the number of MQI calls issued by a queue manager.

Queue statistics messages

Queue statistics messages contain information relating to the activity of a queue during a configured interval. The information includes the number of messages put on, and retrieved from, the queue, and the total number of bytes processed by a queue.

Each queue statistics message can contain up to 100 records, with each record relating to the activity per queue for which statistics were collected.

Statistics messages are recorded only for local queues. If an application makes an MQI call against an alias queue, the statistics data is recorded against the base queue, and, for a remote queue, the statistics data is recorded against the transmission queue.

Channel statistics messages

Channel statistics messages contain information relating to the activity of a channel during a configured interval. For example the information might be the number of messages transferred by the channel, or the number of bytes transferred by the channel.

Each channel statistics message contains up to 100 records, with each record relating to the activity per channel for which statistics were collected.


Statistics messages format

Statistics messages comprise a set of PCF fields that consist of a message descriptor and message data.

Message descriptor

  • A statistics message MQMD (message descriptor)

Accounting message data

  • A statistics message MQCFH (PCF header)
  • Statistics message data that is always returned
  • Statistics message data that is returned if available

The statistics message MQCFH (PCF header) contains information about the interval for which the statistics data was recorded.

Statistics message data comprises PCF parameters that store the statistics information. The content of statistics messages depends on the message category as follows:

MQI statistics message

MQI statistics message data consists of a number of PCF parameters, but no PCF groups.

Queue statistics message

Queue statistics message data consists of a number of PCF parameters, and in the range 1 through 100 QStatisticsData PCF groups.

There is one QStatisticsData PCF group for every queue was active in the interval. If more than 100 queues were active in the interval, multiple statistics messages are generated. Each message has the SeqNumber in the MQCFH (PCF header) updated accordingly, and the last message in the sequence has the Control parameter in the MQCFH specified as MQCFC_LAST.

Channel statistics message

Channel statistics message data consists of a number of PCF parameters, and in the range 1 through 100 ChlStatisticsData PCF groups.

There is one ChlStatisticsData PCF group for every channel that was active in the interval. If more than 100 channels were active in the interval, multiple statistics messages are generated. Each message has the SeqNumber in the MQCFH (PCF header) updated accordingly, and the last message in the sequence has the Control parameter in the MQCFH specified as MQCFC_LAST.


Concept Feedback


Last updated: Monday, September 23, 2013



Statistics information collection

Use queue, queue manager, and channel attributes to control the collection of statistics information


MQI statistics information

Use the queue manager attribute STATMQI to control the collection of MQI statistics information

To change the value of this r attribute, use the MQSC command, ALTER QMGR and specify the parameter STATMQI. Statistics messages are generated only for queues that are opened after statistics collection has been enabled. The STATMQI parameter can have the following values:

ON

MQI statistics information is collected for every connection to the queue manager.

OFF

MQI statistics information is not collected. This is the default value.

For example, to enable MQI statistics information collection use the following MQSC command:

 ALTER QMGR STATMQI(ON)


Reference Feedback


Last updated: Monday, September 23, 2013



Queue statistics information

Use the queue attribute STATQ and the queue manager attribute STATQ to control the collection of queue statistics information

You can enable or disable queue statistics information collection for individual queues or for multiple queues. To control individual queues, set the queue attribute STATQ. You enable or disable queue statistics information collection at the queue manager level by using the queue manager attribute STATQ. For all queues that have the queue attribute STATQ specified with the value QMGR, queue statistics information collection is controlled at the queue manager level.

Queue statistics are incremented only for operations using WebSphere MQ MQI Object Handles that were opened after statistics collection has been enabled.

Queue Statistics messages are generated only for queues for which statistics data has been collected in the previous time period.

The same queue can have several put operations and get operations through several Object Handles. Some Object Handles might have been opened before statistics collection was enabled, but others were opened afterwards. Therefore, it is possible for the queue statistics to record the activity of some put operations and get operations, and not all.

To ensure that the Queue Statistics are recording the activity of all applications, you must close and reopen new Object Handles on the queue, or queues, that you are monitoring. The best way to achieve this, is to end and restart all applications after enabling statistics collection.

To change the value of the queue attribute STATQ, use the MQSC command, ALTER QLOCAL and specify the parameter STATQ. The queue attribute STATQ can have the following values:

ON

Queue statistics information is collected for every connection to the queue manager that opens the queue.

OFF

Queue statistics information for this queue is not collected.

QMGR

The collection of queue statistics information for this queue is controlled according to the value of the queue manager attribute, STATQ. This is the default value.

To change the value of the queue manager attribute STATQ, use the MQSC command, ALTER QMGR and specify the parameter STATQ. The queue manager attribute STATQ can have the following values:

ON

Queue statistics information is collected for queues that have the queue attribute STATQ set as QMGR

OFF

Queue statistics information is not collected for queues that have the queue attribute STATQ set as QMGR. This is the default value.

NONE

The collection of queue statistics information is disabled for all queues, regardless of the queue attribute STATQ.

If the queue manager attribute STATQ is set to NONE, the collection of queue statistics information is disabled for all queues, regardless of the queue attribute STATQ.

For example, to enable statistics information collection for the queue, Q1, use the following MQSC command:

 ALTER QLOCAL(Q1) STATQ(ON)
To enable statistics information collection for all queues that specify the queue attribute STATQ as QMGR, use the following MQSC command:

 ALTER QMGR STATQ(ON)


Reference Feedback


Last updated: Monday, September 23, 2013



Channel statistics information

Use the channel attribute STATCHL to control the collection of channel statistics information. You can also set queue manager attributes to control information collection.

You can enable or disable channel statistics information collection for individual channels, or for multiple channels. To control individual channels, you must set the channel attribute STATCHL to enable or disable channel statistic information collection. To control many channels together, you enable or disable channel statistics information collection at the queue manager level by using the queue manager attribute STATCHL. For all channels that have the channel attribute STATCHL specified with the value QMGR, channel statistics information collection is controlled at the queue manager level.

Automatically defined cluster-sender channels are not WebSphere MQ objects, so do not have attributes in the same way as channel objects. To control automatically defined cluster-sender channels, use the queue manager attribute STATACLS. This attribute determines whether automatically defined cluster-sender channels within a queue manager are enabled or disabled for channel statistics information collection.

You can set channel statistics information collection to one of the three monitoring levels: low, medium or high. You can set the monitoring level at either object level or at the queue manager level. The choice of which level to use is dependant on your system. Collecting statistics information data might require some instructions that are relatively expensive computationally, so to reduce the impact of channel statistics information collection, the medium and low monitoring options measure a sample of the data at regular intervals rather than collecting data all the time. Table 1 summarizes the levels available with channel statistics information collection:

Detail level of channel statistics information collection

Level Description Usage
Low

Measure a small sample of the data, at regular intervals.

For objects that process a high volume of messages.

Medium

Measure a sample of the data, at regular intervals.

For most objects.

High

Measure all data, at regular intervals.

For objects that process only a few messages per second, on which the most current information is important.

To change the value of the channel attribute STATCHL, use the MQSC command, ALTER CHANNEL and specify the parameter STATCHL.

To change the value of the queue manager attribute STATCHL, use the MQSC command, ALTER QMGR and specify the parameter STATCHL.

To change the value of the queue manager attribute STATACLS, use the MQSC command, ALTER QMGR and specify the parameter STATACLS.

The channel attribute, STATCHL, can have the following values:

LOW

Channel statistics information is collected with a low level of detail.

MEDIUM

Channel statistics information is collected with a medium level of detail.

HIGH

Channel statistics information is collected with a high level of detail.

OFF

Channel statistics information is not collected for this channel.

This is the default value.

QMGR

The channel attribute is set as QMGR. The collection of statistics information for this channel is controlled by the value of the queue manager attribute, STATCHL.
The queue manager attribute, STATCHL, can have the following values:

LOW

Channel statistics information is collected with a low level of detail, for all channels that have the channel attribute STATCHL set as QMGR.

MEDIUM

Channel statistics information is collected with a medium level of detail, for all channels that have the channel attribute STATCHL set as QMGR.

HIGH

Channel statistics information is collected with a high level of detail, for all channels that have the channel attribute STATCHL set as QMGR.

OFF

Channel statistics information is not collected for all channels that have the channel attribute STATCHL set as QMGR.

This is the default value.

NONE

The collection of channel statistics information is disabled for all channel, regardless of the channel attribute STATCHL.

The queue manager attribute, STATACLS, can have the following values:

LOW

Statistics information is collected with a low level of detail for automatically defined cluster-sender channels.

MEDIUM

Statistics information is collected with a medium level of detail for automatically defined cluster-sender channels.

HIGH

Statistics information is collected with a high level of detail for automatically defined cluster-sender channels.

OFF

Statistics information is not for automatically defined cluster-sender channels.

This is the default value.

QMGR

The collection of statistics information for automatically defined cluster-sender channels is controlled by the value of the queue manager attribute, STATCHL.

For example, to enable statistics information collection, with a medium level of detail, for the sender channel QM1.TO.QM2, use the following MQSC command:

 ALTER CHANNEL(QM1.TO.QM2) CHLTYPE(SDR) STATCHL(MEDIUM) 
To enable statistics information collection, at a medium level of detail, for all channels that specify the channel attribute STATCHL as QMGR, use the following MQSC command:

 ALTER QMGR STATCHL(MEDIUM)
To enable statistics information collection, at a medium level of detail, for all automatically defined cluster-sender channels, use the following MQSC command:

 ALTER QMGR STATACLS(MEDIUM)


Reference Feedback


Last updated: Monday, September 23, 2013



Statistics message generation

Statistics messages are generated at configured intervals, and when a queue manager shuts down in a controlled fashion.

The configured interval is controlled by the STATINT queue manager attribute, which specifies the interval, in seconds, between the generation of statistics messages. The default statistics interval is 1800 seconds (30 minutes). To change the statistics interval, use the MQSC command ALTER QMGR and specify the STATINT parameter. For example, to change the statistics interval to 900 seconds (15 minutes) use the following MQSC command:

 ALTER QMGR STATINT(900)

To write the currently collected statistics data to the statistics queue before the statistics collection interval is due to expire, use the MQSC command RESET QMGR TYPE(STATISTICS). Issuing this command causes the collected statistics data to be written to the statistics queue and a new statistics data collection interval to begin.


Concept Feedback


Last updated: Monday, September 23, 2013



Display accounting and statistics information

To use the information recorded in accounting and statistics messages, run an application such as the amqsmon sample program to transform the recorded information into a suitable format

Accounting and statistics messages are written to the system accounting and statistics queues. amqsmon is a sample program supplied with WebSphere MQ that processes messages from the accounting and statistics queues and displays the information to the screen in a readable form.

Because amqsmon is a sample program, you can use the supplied source code as template for writing your own application to process accounting or statistics messages, or modify the amqsmon source code to meet your own particular requirements.


amqsmon (Display formatted monitoring information)

Use the amqsmon sample program to display in a readable format the information contained within accounting and statistics messages. The amqsmon program reads accounting messages from the accounting queue, SYSTEM.ADMIN.ACCOUNTING.QUEUE. and reads statistics messages from the statistics queue, SYSTEM.ADMIN.STATISTICS.QUEUE.


Syntax

   
>>-amqsmon--+----------------+---------------------------------->
            '- -m -- QMgrName-'   

>-- -t -- Type--+-----------------------+--+------+-------------->
               +- -a ------------------+  '- -b -'   
               +- -i -- ConnectionId----+             
               +- -c --+-------------+-+             
               |       '- ChannelName-' |             
               '- -q --+-----------+---'             
                       '- QueueName-'                 

>--+-------------+--+---------------+--+-----------------+------>
   '- -d -- Depth-'  '- -w -- TimeOut-'  '- -s -- StartTime-'   

>--+---------------+--+-------------------------+--------------><
   '- -e -- EndTime-'  |       .-,-------------. |   
                      |       V               | |   
                      '- -l ----+-----------+-+-'   
                                '- Parameter-'       



Required parameters

-t Type
The type of messages to process. Specify Type as one of the following:

accounting

Accounting records are processed. Messages are read from the system queue, SYSTEM.ADMIN.ACCOUNTING.QUEUE.

statistics

Statistics records are processed. Messages are read from the system queue, SYSTEM.ADMIN.STATISTICS.QUEUE.


Optional Parameters

-m QMgrName
The name of the queue manager from which accounting or statistics messages are to be processed.

If you do not specify this parameter, the default queue manager is used.

-a
Process messages containing MQI records only.

Only display MQI records. Messages not containing MQI records will always be left on the queue they were read from.

-q QueueName
QueueName is an optional parameter.

-c ChannelName
ChannelName is an optional parameter.

This parameter is available when displaying statistics messages only, (-t statistics).

-i ConnectionId
Displays records related to the connection identifier specified by ConnectionId only.

This parameter is available when displaying accounting messages only, (-t accounting).

If -b is not specified then the statistics messages from which the records came are discarded. Since statistics messages can also contain records from other channels, if -b is not specified then unseen records can be discarded.

-b
Browse messages.

Messages are retrieved non-destructively.

-d Depth
The maximum number of messages that can be processed.

If you do not specify this parameter, then an unlimited number of messages can be processed.

-w TimeOut
Time maximum number of seconds to wait for a message to become available.

If you do not specify this parameter, amqsmon will end once there are no more messages to process.

-s StartTime
Process messages put after the specified StartTime only.

StartTime is specified in the format yyyy-mm-dd hh.mm.ss. If a date is specified without a time, then the time will default to 00.00.00 on the date specified. Times are in GMT.

For the effect of not specifying this parameter, see Note 1 .

-e EndTime
Process messages put before the specified EndTime only.

The EndTime is specified in the format yyyy-mm-dd hh.mm.ss. If a date is specified without a time, then the time will default to 23.59.59 on the date specified. Times are in GMT.

For the effect of not specifying this parameter, see Note 1 .

-l Parameter
Only display the selected fields from the records processed. Parameter is a comma-separated list of integer values, with each integer value mapping to the numeric constant of a field, see amqsmon example 5 .

If you do not specify this parameter, then all available fields are displayed.

Note:

  1. If you do not specify -s StartTime or -e EndTime, the messages that can be processed are not restricted by put time.


Reference Feedback


Last updated: Monday, September 23, 2013



amqsmon examples

Use this page to view examples of running the amqsmon (Display formatted monitoring information) sample program

  1. The following command displays all MQI statistics messages from queue manager saturn.queue.manager:

     amqsmon -m saturn.queue.manager -t statistics -a
    
    The output from this command follows:

       RecordType: MQIStatistics
       QueueManager: 'saturn.queue.manager'
       IntervalStartDate: '2005-04-30'
       IntervalStartTime: '15.09.02'
       IntervalEndDate: '2005-04-30'
       IntervalEndTime: '15.39.02'
       CommandLevel: 600
       ConnCount: 23
       ConnFailCount: 0
       ConnHighwater: 8
       DiscCount: [17, 0, 0]
       OpenCount: [0, 80, 1, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0]
       OpenFailCount: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       CloseCount: [0, 73, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0]
       CloseFailCount: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       InqCount: [4, 2102, 0, 0, 0, 46, 0, 0, 0, 0, 0, 0, 0]
       InqFailCount: [0, 31, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       SetCount: [0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0]
       SetFailCount: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       PutCount: [26, 1]
       PutFailCount: 0
       Put1Count: [40, 0]
       Put1FailCount: 0
       PutBytes: [57064, 12320]
       GetCount: [18, 1]
       GetBytes: [52, 12320]
       GetFailCount: 2254
       BrowseCount: [18, 60]
       BrowseBytes: [23784, 30760]
       BrowseFailCount: 9
       CommitCount: 0
       CommitFailCount: 0
       BackCount: 0
       ExpiredMsgCount: 0
       PurgeCount: 0
    

  2. The following command displays all queue statistics messages for queue LOCALQ on queue manager saturn.queue.manager:

     amqsmon -m saturn.queue.manager -t statistics -q LOCALQ
    
    The output from this command follows:

       RecordType: QueueStatistics
       QueueManager: 'saturn.queue.manager'
       IntervalStartDate: '2005-04-30'
       IntervalStartTime: '15.09.02'
       IntervalEndDate: '2005-04-30'
       IntervalEndTime: '15.39.02'
       CommandLevel: 600
       ObjectCount: 3
       QueueStatistics:
         QueueName: 'LOCALQ'
         CreateDate: '2005-03-08'
         CreateTime: '17.07.02'
         QueueType: Predefined
         QueueDefinitionType: Local
         QMinDepth: 0
         QMaxDepth: 18
         AverageQueueTime: [29827281, 0]
         PutCount: [26, 0]
         PutFailCount: 0
         Put1Count: [0, 0]
         Put1FailCount: 0
         PutBytes: [88, 0]
         GetCount: [18, 0]
         GetBytes: [52, 0]
         GetFailCount: 0
         BrowseCount: [0, 0]
         BrowseBytes: [0, 0]
         BrowseFailCount: 1
         NonQueuedMsgCount: 0
         ExpiredMsgCount: 0
         PurgedMsgCount: 0
    
  3. The following command displays all of the statistics messages recorded since 15:30 on 30 April 2005 from queue manager saturn.queue.manager.

     amqsmon -m saturn.queue.manager -t statistics -s "2005-04-30 15.30.00"
    
    The output from this command follows:

       RecordType: MQIStatistics
       QueueManager: 'saturn.queue.manager'
       IntervalStartDate: '2005-04-30'
       IntervalStartTime: '15.09.02'
       IntervalEndDate: '2005-04-30'
       IntervalEndTime: '15.39.02'
       CommandLevel: 600
       ConnCount: 23
       ConnFailCount: 0
       ConnHighwater: 8
       DiscCount: [17, 0, 0]
       OpenCount: [0, 80, 1, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0]
         ... 
       RecordType: QueueStatistics
       QueueManager: 'saturn.queue.manager'
       IntervalStartDate: '2005-04-30'
       IntervalStartTime: '15.09.02'
       IntervalEndDate: '2005-04-30'
       IntervalEndTime: '15.39.02'
       CommandLevel: 600
       ObjectCount: 3
       QueueStatistics: 0
         QueueName: 'LOCALQ'
         CreateDate: '2005-03-08'
         CreateTime: '17.07.02'
         QueueType: Predefined
           ... 
       QueueStatistics: 1
         QueueName: 'SAMPLEQ'
         CreateDate: '2005-03-08'
         CreateTime: '17.07.02'
         QueueType: Predefined
           ... 
    
  4. The following command displays all accounting messages recorded on 30 April 2005 from queue manager saturn.queue.manager:

     amqsmon -m saturn.queue.manager -t accounting -s "2005-04-30" -e "2005-04-30"
    
    The output from this command follows:

       RecordType: MQIAccounting
       QueueManager: 'saturn.queue.manager'
       IntervalStartDate: '2005-04-30'
       IntervalStartTime: '15.09.29'
       IntervalEndDate: '2005-04-30'
       IntervalEndTime: '15.09.30'
       CommandLevel: 600
       ConnectionId: x'414d51435452455631202020202020208d0b3742010a0020'
       SeqNumber: 0
       ApplicationName: 'amqsput'
       ApplicationPid: 8572
       ApplicationTid: 1
       UserId: 'admin'
       ConnDate: '2005-03-16'
       ConnTime: '15.09.29'
       DiscDate: '2005-03-16'
       DiscTime: '15.09.30'
       DiscType: Normal
       OpenCount: [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       OpenFailCount: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       CloseCount: [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       CloseFailCount: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       PutCount: [1, 0]
       PutFailCount: 0
       PutBytes: [4, 0]
       GetCount: [0, 0]
       GetFailCount: 0
       GetBytes: [0, 0]
       BrowseCount: [0, 0]
       BrowseFailCount: 0
       BrowseBytes: [0, 0]
       CommitCount: 0
       CommitFailCount: 0
       BackCount: 0
       InqCount: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       InqFailCount: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       SetCount: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       SetFailCount: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
       
       RecordType: MQIAccounting
       QueueManager: 'saturn.queue.manager'
       IntervalStartDate: '2005-03-16'
       IntervalStartTime: '15.16.22'
       IntervalEndDate: '2005-03-16'
       IntervalEndTime: '15.16.22'
       CommandLevel: 600
       ConnectionId: x'414d51435452455631202020202020208d0b3742010c0020'
       SeqNumber: 0
       ApplicationName: 'runmqsc'
       ApplicationPid: 8615
       ApplicationTid: 1
           ... 
    
  5. The following command browses the accounting queue and displays the application name and connection identifier of every application for which MQI accounting information is available:

     amqsmon -m saturn.queue.manager -t accounting -b -a -l 7006,3024
    
    The output from this command follows:

       ConnectionId: x'414d51435452455631202020202020208d0b374203090020'
       ApplicationName: 'runmqsc'
    
       ConnectionId: x'414d51435452455631202020202020208d0b3742010a0020'
       ApplicationName: 'amqsput'
    
       ConnectionId: x'414d51435452455631202020202020208d0b3742010c0020'
       ApplicationName: 'runmqsc'
    
       ConnectionId: x'414d51435452455631202020202020208d0b3742010d0020'
       ApplicationName: 'amqsput'
    
       ConnectionId: x'414d51435452455631202020202020208d0b3742150d0020'
       ApplicationName: 'amqsget'
    
       5 Records Processed.
    


Reference Feedback


Last updated: Monday, September 23, 2013



Accounting and statistics message reference

Use this page to obtain an overview of the format of accounting and statistics messages and the information returned in these messages

Accounting and statistics message messages are standard WebSphere MQ messages containing a message descriptor and message data. The message data contains information about the MQI operations performed by WebSphere MQ applications, or information about the activities occurring in a WebSphere MQ system.

Message descriptor

  • An MQMD structure

Message data

  • A PCF header (MQCFH)
  • Accounting or statistics message data that is always returned
  • Accounting or statistics message data that is returned if available


Accounting and statistics message format

Use this page as an example of the structure of an MQI accounting message

MQI accounting message structure

MQMD structure Accounting message header MQCFH structure MQI accounting message data 1

 
Structure identifier
Structure version
Report options
Message type
Expiration time
Feedback code
Encoding
Coded character set ID
Message format
Message priority
Persistence
Message identifier
Correlation identifier
Backout count
Reply-to queue
Reply-to queue manager
User identifier
Accounting token
Application identity data
Application type
Application name
Put date
Put time
Application origin data
Group identifier
Message sequence number
Offset
Message flags
Original length
 

 
Structure type
Structure length
Structure version
Command identifier
Message sequence number
Control options
Completion code
Reason code
Parameter count
 

 
Queue manager
Interval start date
Interval start time
Interval end date
Interval end time
Command level
Connection identifier
Sequence number
Application name
Application process identifier
Application thread identifier
User identifier
Connection date
Connection time
Connection name
Channel name
Disconnect date
Disconnect time
Disconnect type
Open count
Open fail count
Close count
Close fail count
Put count
Put fail count
Put1 count
Put1 fail count
Put bytes
Get count
Get fail count
Get bytes
Browse count
Browse fail count
Browse bytes
Commit count
Commit fail count
Backout count
Inquire count
Inquire fail count
Set count
Set fail count
 

Note:

  1. The parameters shown are those returned for an MQI accounting message. The actual accounting or statistics message data depends on the message category.


Concept Feedback


Last updated: Monday, September 23, 2013



Accounting and statistics message MQMD (message descriptor)

Use this page to understand the differences between the message descriptor of accounting and statistics messages and the message descriptor of event messages

The parameters and values in the message descriptor of accounting and statistics message are the same as in the message descriptor of event messages, with the following exception:

Format

Some of the parameters contained in the message descriptor of accounting and statistics message contain fixed data supplied by the queue manager that generated the message.

The MQMD also specifies the name of the queue manager (truncated to 28 characters) that put the message, and the date and time when the message was put on the accounting, or statistics, queue.


Reference Feedback


Last updated: Monday, September 23, 2013



Message data in accounting and statistics messages

The message data in accounting and statistics messages is based on the programmable command format (PCF), which is used in PCF command inquiries and responses. The message data in accounting and statistics messages consists of a PCF header (MQCFH) and an accounting or statistics report.


Accounting and statistics message MQCFH (PCF header)

The message header of accounting and statistics messages is an MQCFH structure. The parameters and values in the message header of accounting and statistics message are the same as in the message header of event messages, with the following exceptions:

Command

Version


Accounting and statistics message data

The content of accounting and statistics message data is dependent on the category of the accounting or statistics message, as follows:

MQI accounting message

MQI accounting message data consists of a number of PCF parameters, but no PCF groups.

Queue accounting message

Queue accounting message data consists of a number of PCF parameters, and in the range 1 through 100 QAccountingData PCF groups.

MQI statistics message

MQI statistics message data consists of a number of PCF parameters, but no PCF groups.

Queue statistics message

Queue statistics message data consists of a number of PCF parameters, and in the range 1 through 100 QStatisticsData PCF groups.

Channel statistics message

Channel statistics message data consists of a number of PCF parameters, and in the range 1 through 100 ChlStatisticsData PCF groups.


Concept Feedback


Last updated: Monday, September 23, 2013



MQI accounting message data

Use this page to view the structure of an MQI accounting message

QueueManager

IntervalStartDate

IntervalStartTime

IntervalEndDate

IntervalEndTime

CommandLevel

ConnectionId

SeqNumber

ApplicationName

ApplicationPid

ApplicationTid

UserId

ConnDate

ConnTime

ConnName

ChannelName

DiscDate

DiscTime

DiscType

OpenCount

OpenFailCount

CloseCount

CloseFailCount

PutCount

PutFailCount

Put1Count

Put1FailCount

PutBytes

GetCount

GetFailCount

GetBytes

BrowseCount

BrowseFailCount

BrowseBytes

CommitCount

CommitFailCount

BackCount

InqCount

InqFailCount

SetCount

SetFailCount

SubCountDur

SubCountNDur

SubFailCount

UnsubCountDur

UnsubCountNDur

UnsubFailCount

SubRqCount

SubRqFailCount

CBCount

CBFailCount

CtlCount

CtlFailCount

StatCount

StatFailCount

PutTopicCount

PutTopicFailCount

Put1TopicCount

Put1TopicFailCount

PutTopicBytes


Reference Feedback


Last updated: Monday, September 23, 2013



Queue accounting message data

Use this page to view the structure of a queue accounting message

QueueManager

IntervalStartDate

IntervalStartTime

IntervalEndDate

IntervalEndTime

CommandLevel

ConnectionId

SeqNumber

ApplicationName

ApplicationPid

ApplicationTid

UserId

ObjectCount

QAccountingData

QName

CreateDate

CreateTime

QType

QDefinitionType

OpenCount

OpenDate

OpenTime

CloseDate

CloseTime

PutCount

PutFailCount

Put1Count

Put1FailCount

PutBytes

PutMinBytes

PutMaxBytes

GeneratedMsgCount

GetCount

GetFailCount

GetBytes

GetMinBytes

GetMaxBytes

BrowseCount

BrowseFailCount

BrowseBytes

BrowseMinBytes

BrowseMaxBytes

CBCount

CBFailCount

TimeOnQMin

TimeOnQAvg

TimeOnQMax


Reference Feedback


Last updated: Monday, September 23, 2013



MQI statistics message data

Use this page to view the structure of an MQI statistics message

QueueManager

IntervalStartDate

IntervalStartTime

IntervalEndDate

IntervalEndTime

CommandLevel

ConnCount

ConnFailCount

ConnsMax

DiscCount

OpenCount

OpenFailCount

CloseCount

CloseFailCount

InqCount

InqFailCount

SetCount

SetFailCount

PutCount

PutFailCount

Put1Count

Put1FailCount

PutBytes

GetCount

GetFailCount

GetBytes

BrowseCount

BrowseFailCount

BrowseBytes

CommitCount

CommitFailCount

BackCount

ExpiredMsgCount

PurgeCount

SubCountDur

SubCountNDur

SubFailCount

UnsubCountDur

UnsubCountNDur

UnsubFailCount

SubRqCount

SubRqFailCount

CBCount

CBFailCount

CtlCount

CtlFailCount

StatCount

StatFailCount

SubCountDurHighWater

SubCountDurLowWater

SubCountNDurHighWater

SubCountNDurLowWater

PutTopicCount

PutTopicFailCount

Put1TopicCount

Put1TopicFailCount

PutTopicBytes

PublishMsgCount

PublishMsgBytes


Reference Feedback


Last updated: Monday, September 23, 2013



Queue statistics message data

Use this page to view the structure of a queue statistics message

QueueManager

IntervalStartDate

IntervalStartTime

IntervalEndDate

IntervalEndTime

CommandLevel

ObjectCount

QStatisticsData

QName

CreateDate

CreateTime

QType

QDefinitionType

QMinDepth

QMaxDepth

AvgTimeOnQ

PutCount

PutFailCount

Put1Count

Put1FailCount

PutBytes

GetCount

GeneratedMsgCount

GetFailCount

GetBytes

BrowseCount

BrowseFailCount

BrowseBytes

NonQueuedMsgCount

ExpiredMsgCount

PurgeCount

CBCount

CBFailCount


Reference Feedback


Last updated: Monday, September 23, 2013



Channel statistics message data

Use this page to view the structure of a channel statistics message

QueueManager

IntervalStartDate

IntervalStartTime

IntervalEndDate

IntervalEndTime

CommandLevel

ObjectCount

ChlStatisticsData

ChannelName

ChannelType

RemoteQmgr

ConnectionName

MsgCount

TotalBytes

NetTimeMin

NetTimeAvg

NetTimeMax

ExitTimeMin

ExitTimeAvg

ExitTimeMax

FullBatchCount

IncmplBatchCount

AverageBatchSize

PutRetryCount


Reference Feedback


Last updated: Monday, September 23, 2013



Reference notes

Use this page to view the notes to which descriptions of the structure of accounting and statistics messages refer

The following message data descriptions refer to these notes:

  1. This parameter relates to WebSphere MQ objects. This parameter is an array of values (MQCFIL or MQCFIL64) indexed by the following constants:

    Array indexed by object type

    Object type Value context
    MQOT_Q (1) Contains the value relating to queue objects.
    MQOT_NAMELIST (2) Contains the value relating to namelist objects.
    MQOT_PROCESS (3) Contains the value relating to process objects.
    MQOT_Q_MGR (5) Contains the value relating to queue manager objects.
    MQOT_CHANNEL (6) Contains the value relating to channel objects.
    MQOT_AUTH_INFO (7) Contains the value relating to authentication information objects.
    MQOT_TOPIC (8) Contains the value relating to topic objects.

    Note: An array of 13 MQCFIL or MQCFIL64 values are returned but only those listed are meaningful.

  2. This parameter relates to WebSphere MQ messages. This parameter is an array of values (MQCFIL or MQCFIL64) indexed by the following constants:

    Array indexed by persistence value

    Constant Value
    1 Contains the value for nonpersistent messages.
    2 Contains the value for persistent messages.

Note: The index for each of these arrays starts at zero, so an index of 1 refers to the second row of the array. Elements of these arrays not listed in these tables contain no accounting or statistics information.


Reference Feedback


Last updated: Monday, September 23, 2013



Application activity trace

Application activity trace produces detailed information about the behavior of applications connected to a queue manager. It traces the behavior of an application and provides a detailed view of the parameters used by an application as it interacts with WebSphere MQ resources. It also shows the sequence of MQI calls issued by an application.

Use Application activity trace when you require more information than what is provided by Event monitoring, Message monitoring, Accounting and statistics messages, and Real-time monitoring.


Application activity trace messages

Application activity trace produces detailed information about the behavior of applications connected to a queue manager. An application activity trace message is a PCF message that contains a number of PCF structures.


Application activity trace information collection

Use queue and queue manager attributes to control the collection of application activity trace information. You can also use MQCONNX options to control collection at the connection level.

Activity records are written to the new system queue SYSTEM.ADMIN.TRACE.ACTIVITY.QUEUE. They collect information about the MQI operations that an application performs while connected to a queue manager. Messages are normally written when the application disconnects from the queue manager. For long running applications, intermediate messages will be written if the lifetime of the connection exceeds a user-defined timeout, or after a user-defined number of operations, or when the size of the record approaches the largest message allowed on the trace queue.

The timeout is defined by the ActivityInterval parameter, and the count is defined by the ActivityCount parameter. Both of which can be specified in the activity trace configuration file.

When the timeout has expired or MQI count is reached then the activity data that has been accumulated so far is written as a message and the activity data is reset.

Intermediate messages are also written when the amount of data collected in memory reaches the maximum message length allowed for the SYSTEM.ADMIN.TRACE.ACTIVITY.QUEUE.

The activity trace messages are composed of an MQMD structure, a PCF (MQCFH) header structure followed by a number of PCF parameters. A sequence of ApplicationTraceData PCF groups describing the MQI activity of the application follows the PCF parameters.

Trace entries are added after each operation has completed unless otherwise stated.


MQI application activity trace information

Use the queue manager attribute ACTVTRC to control the collection of MQI application activity trace information

To change the value of this attribute, use the MQSC command, ALTER QMGR, and specify the parameter ACTVTRC. Application activity trace messages are generated only for connections that begin after application activity trace is enabled. The ACTVTRC parameter can have the following values:

ON

API activity trace collection is switched on.

OFF

API activity trace collection is switched off.

For example, to enable MQI application activity trace information collection use the following MQSC command:

 ALTER QMGR ACTVTRC(ON)


Concept Feedback


Last updated: Monday, September 23, 2013



MQCONNX options

Use the ConnectOpts parameter on the MQCONNX call to enable or disable application activity reports on a per connection basis. If neither of these options is used then the activity trace behavior is defined by the Queue manager attribute ACTVTRC

The ConnectOpts parameter can have the following values:

MQCNO_ACTIVITY_ TRACE_DISABLED

Activity trace is turned off for the connection.

MQCNO_ ACTIVITY_ TRACE_ENABLED

Activity trace is turned on for the connection.

If an application selects both ENABLED and DISABLED for MQCONNX, the call fails with a reason code of MQRC_OPTIONS_ERROR.

If an application attempts to modify the accounting behavior of an application using these flags but the QMGR attribute ACTVCONO is set to DISABLED then no error is returned to the application and activity trace collection is defined by the Qmgr attributes and the activity trace ini file.

MQCONNX values might be ignored based on the ACTVCONO value.


Concept Feedback


Last updated: Monday, September 23, 2013



Activity Trace configuration file

The Activity Trace behavior is configured using a configuration file which follows the same stanza key and parameter-value pair format as the mqs.ini and qm.ini files. The activity trace configuration file is called mqat.ini. On UNIX and Linux systems, mqat.ini is located in the queue manager data directory (the same location as the qm.ini file). On Windows systems, mqat.ini is located in the queue manager data directory, C:\Program Files\IBM\WebSphere MQ\qmgrs. Users running applications to be traced require permission to read this file.

The syntax rules for the format of the file are:


Stanza keys

Two stanza key types are allowed in the configuration file: the AllActivityTrace stanza, and the ApplicationTrace stanza


AllActivityTrace stanza

The AllActivityTrace stanza defines settings for the activity trace that is applied to all WebSphere MQ connections unless overridden.

Individual values in the AllActivityTrace stanza can be overridden by more specific information in an ApplicationTrace stanza.

If more than one AllActivityTrace stanza is specified then the values in the last stanza is used. Parameters missing from the chosen AllActivityTrace take default values. Parameters and values from previous AllActivityTrace stanzas are ignored


ApplicationTrace stanza

The ApplicationTrace stanza defines settings which can be applied to a specific name, type or both of WebSphere MQ connection.

This stanza includes ApplName and ApplClass values which are used according to the matching rules defined in Connection Matching Rules to determine whether the stanza applies to a particular connection.


Parameter/Value Pairs

The following table lists the parameter/value pairs which may be used in the activity trace configuration file.

Parameter/value pairs that can be used in the activity trace configuration file

Name Stanza Type Values (default in bold type) Description
Trace ApplicationTrace ON / OFF Activity trace switch. This switch can be used in the application-specific stanza to determine whether activity trace is active for the scope of the current application stanza.
ActivityInterval AllActivityTrace ApplicationTrace 0 -99999999 ( 0=off ) Time interval in seconds between trace messages. Activity trace does not use a timer thread, so the trace message will not be written at the exact instant that the time elapses - rather it will be written when the first MQI operation is executed after the time interval has elapsed. If this value is 0 then the trace message is written when the connection disconnects (or when the activity count is reached).
ActivityCount AllActivityTrace ApplicationTrace 0 -99999999 ( 0=off ) Number of MQI or XA operations between trace messages. If this value is 0 then the trace message is written when the connection disconnects (or when the activity interval has elapsed).
TraceLevel AllActivityTrace ApplicationTrace LOW / MEDIUM / HIGH Amount of parameter detail traced for each operation. The description of individual operations details which parameters are included for each trace level.
TraceMessageData AllActivityTrace ApplicationTrace 0 - 104 857 600 (100Mb) Amount of message data traced in bytes for MQGET, MQPUT, MQPUT1, and Callback operations
ApplName ApplicationTrace Character string (Required parameter - no default) This value is used to determine which applications the ApplicationTrace stanza applies to. It is matched to the ApplName value from the API exit context structure (which is equivalent to the MQMD.PutApplName). The content of the ApplName value varies according to the application environment. For distributed platforms, only the filename portion of the MQAXC.ApplName is matched to the value in the stanza. Characters to the left of the rightmost path separator are ignored when the comparison is made. For z/OS applications, the entire MQAXC.ApplName is matched to the value in the stanza. A single wildcard character (*) can be used at the end of the ApplName value to match any number of characters after that point are. If the ApplName value is set to a single wildcard character (*) then the ApplName value matches all applications.
ApplClass ApplicationTrace USER / MCA / INTERNAL / ALL The class of application. See the following table for an explanation of how the AppType values correspond to WebSphere MQ connections

The following table shows how the AppClass values correspond to the APICallerType and APIEnvironment fields in the connection API exit context structure.

Appclass values and how they correspond to the APICallerType and APIEnvironment fields

APPLCLASS API Caller Type: API Environment: Description
USER MQXACT_EXTERNAL MQXE_OTHER Only user applications are traced
MCA (Any value) MQXE_MCA
MQXE_MCA_CLNTCONN
MQXE_MCA_SVRCONN
Clients and channels (amqrmppa)
INTERNAL MQXACT_EXTERNAL MQXE_COMMAND_SERVER
MQXE_MQSC
'runmqsc' and command server
INTERNAL MQXACT_INTERNAL (Any value) "trusted" and internal applications and processes; for example, amqzdmaa
ALL (Any value) (Any value) All user and internal connections are traced


Connection Matching Rules

The queue manager applies the following rules to determine which stanzas settings to use for a connection.

  1. A value specified in the AllActivityTrace stanza is used for the connection unless the value also occurs in an ApplicationTrace stanza and the stanza fulfills the matching criteria for the connection described in points 2 , 3 , and 4 .
  2. The ApplClass is matched against the type of the WebSphere MQ connection. If the ApplClass does not match the connection type then the stanza is ignored for this connection.
  3. The ApplName value in the stanza is matched against the file name portion of the ApplName field from the API exit context structure (MQAXC) for the connection. The file name portion is derived from the characters to the right of the final path separator (/ or \) character. If the stanza ApplName includes a wildcard (*) then only the characters to the left of the wildcard are compared with the equivalent number of characters from the connections ApplName. For example, if a stanza value of "FRE*" is specified then only the first three characters are used in the comparison, so "path/FREEDOM" and "path\FREDDY" match, but "path/FRIEND" does not. If the stanzas ApplName value does not match the connection ApplName then the stanza is ignored for this connection.
  4. If more than one stanza matches the connections ApplName and ApplClass, then the stanza with the most specific ApplName is used. The most specific ApplName is defined as the one which uses the most characters to match the connections ApplName. For example, if the ini file contains a stanza with ApplName="FRE*" and another stanza with ApplName="FREE*" then the stanza with ApplName="FREE*" is chosen as the best match for a connection with ApplName="path/FREEDOM" because it matches four characters (whereas ApplName="FRE*" matches only three).
  5. If after applying the rules in points 2 , 3 , and 4 , there is more than one stanza that matches the connections ApplName and ApplClass, then the values from the last matching will be used and all other stanzas will be ignored.


Exit Configuration File Example

The following example shows how the configuration data is specified in the Activity Trace ini file. This example is shipped as a sample called mqat.ini in the C samples directory (the same directory as the amqsact.c file)

 AllActivityTrace:
  ActivityInterval=0                    # Time interval between trace messages 
                                        #   Values: 0-99999999 (0=off)
                                        #   Default: 0
  ActivityCount=0                       # Number of operations between trace msgs
                                        #   Values: 0-99999999  (0=off)
                                        #   Default: 0
  TraceLevel=MEDIUM                     # Amount of data traced for each operation
                                        #   Values: LOW | MEDIUM | HIGH 
                                        #   Default: MEDIUM
  TraceMessageData=0                    # Amount of message data traced 
                                        #   Values: 0-100000000
                                        #   Default: 0

ApplicationTrace:
  ApplClass=USER                         # Application type 
                                        #   Values:  (USER | MCA | INTERNAL | ALL)
                                        #   Default: USER
  ApplName=AppName*                     # Application name (may be wildcarded)
                                        #   (matched to app name without path)
                                        #   Default: *
  Trace=OFF                             # Activity trace switch for application 
                                        #   Values:  ( ON | OFF )
                                        #   Default: OFF
ActivityInterval=0                      # Time interval between trace messages 
                                        #   Values: 0-99999999 (0=off)
                                        #   Default: 0
  ActivityCount=0                       # Number of operations between trace msgs
                                        #   Values: 0-99999999  (0=off)
                                        #   Default: 0
  TraceLevel=MEDIUM                     # Amount of data traced for each operation
                                        #   Values: LOW | MEDIUM | HIGH 
                                        #   Default: MEDIUM
  TraceMessageData=0                    # Amount of message data traced 
                                        #   Values: 0-100000000
                                        #   Default: 0


Concept Feedback


Last updated: Monday, September 23, 2013



Performance impact of application activity trace

Enabling application activity trace can incur a performance penalty that can be reduced by tuning ActivityCount and ActivityInterval.

Enabling application activity trace selectively for an application or for all queue manager applications can result in the queue manager requiring additional storage resource and can result in additional messaging activity. In environments where messaging performance is critical, for example, in high workload applications or where a service level agreement (SLA) requires a minimum response time from the messaging provider, it might not be appropriate to collect application activity trace or it might be necessary to adjust the detail or frequency that trace activity messages are produced. The default values of ActivityInterval, ActivityCount and TraceLevel in mqat.ini are configured to give a balance of detail and performance, however it might be necessary to adjust these values to fulfill functional or performance requirements.

To reduce the performance impact incurred when collecting application activity trace, ensure that:


Concept Feedback


Last updated: Monday, September 23, 2013



Application activity trace message reference

Use this page to obtain an overview of the format of application activity trace messages and the information returned in these messages

Application activity trace messages are standard WebSphere MQ messages containing a message descriptor and message data. The message data contains information about the MQI operations performed by WebSphere MQ applications, or information about the activities occurring in a WebSphere MQ system.

Message descriptor

  • An MQMD structure

Message data

  • A PCF header (MQCFH)
  • Application activity trace message data that is always returned
  • Application activity trace message data that is operation-specific


Application activity trace message MQMD (message descriptor)

Use this page to understand the differences between the message descriptor of application activity trace messages and the message descriptor of event messages

The parameters and values in the message descriptor of application activity trace message are the same as in the message descriptor of event messages, with the following exception:

Format

CorrelId


Reference Feedback


Last updated: Monday, September 23, 2013



MQCFH (PCF Header)

Use this page to view the PCF values contained by the MQCFH structure for an activity trace message

For an activity trace message, the MQCFH structure contains the following values:

Type

StrucLength

Version

Command

MsgSeqNumber

Control

CompCode

Reason

ParameterCount


Reference Feedback


Last updated: Monday, September 23, 2013



Application activity trace message data

Immediately following the PCF header is a set of parameters describing the time interval for the activity trace. These parameters also indicate the sequence of messages in the event of messages being written. The order and number of fields following the header is not guaranteed, allowing additional information to be added in the future.

QueueManager

QSGName

HostName

IntervalStartDate

IntervalStartTime

IntervalEndDate

IntervalEndTime

CommandLevel

SeqNumber

ApplicationName

ApplClass

ApplicationPid

UserId

APICallerType

Environment

Detail

TraceDataLength

Pointer Size

Platform


Reference Feedback


Last updated: Monday, September 23, 2013



Variable parameters for application activity MQI operations

The application activity data MQCFGR structure is followed by the set of PCF parameters which corresponds to the operation being performed . The parameters for each operation are defined in the following section.

The trace level indicates the level of trace granularity that is required for the parameters to be included in the trace. The possible trace level values are:

  1. Low

    The parameter is included when low , medium or high activity tracing is configured for an application. This setting means that a parameter is always included in the AppActivityData group for the operation. This set of parameters is sufficient to trace the MQI calls an application makes, and to see if they are successful.

  2. Medium

    The parameter is only included in the AppActivityData group for the operation when medium or high activity tracing is configured for an application. This set of parameters adds information about the resources, for example, queue and topic names used by the application.

  3. High

    The parameter is only included in the AppActivityData group for the operation when high activity tracing is configured for an application. This set of parameters includes memory dumps of the structures passed to the MQI and XA functions. For this reason, it contains more information about the parameters used in MQI and XA calls. The structure memory dumps are shallow copies of the structures. To avoid erroneous attempts to dereference pointers, the pointer values in the structures are set to NULL.

    Note: The version of the structure that is dumped is not necessarily identical to the version used by an application. The structure can be modified by an API crossing exit, by the activity trace code, or by the queue manager. A queue manager can modify a structure to a later version, but the queue manager never changes it to an earlier version of the structure. To do so, would risk losing data.


MQBACK

Application has started the MQBACK MQI function

CompCode

Reason


Reference Feedback


Last updated: Monday, September 23, 2013



MQBEGIN

Application has started the MQBEGIN MQI function

CompCode

Reason

MQBO


Reference Feedback


Last updated: Monday, September 23, 2013



MQCALLBACK

Application has started the MQCALLBACK function

ObjectHandle

CallType

MsgBuffer

MsgLength

HighResTime

ReportOptions

MsgType

Expiry

Format

Priority

Persistence

MsgId

CorrelId

ObjectName

ResolvedQName

ReplyToQueue

ReplyToQMgr

CodedCharSetId

Encoding

PutDate

PutTime

ResolvedQName

ResObjectString

ResolvedType

PolicyName

XmitqMsgId

XmitqCorrelId

XmitqPutTime

XmitqPutDate

XmitqRemoteQName

XmitqRemoteQMgr

MsgDescStructure

GetMsgOptsStructure

MQCBContextStructure


Reference Feedback


Last updated: Monday, September 23, 2013



MQCB

Application has started the manage callback MQI function

CallbackOperation

CallbackType

CallbackOptions

CallbackFunction

CallbackName

ObjectHandle

MaxMsgLength

CompCode

Reason

ResolvedQName

ResObjectString

ResolvedType

CallBack DescriptorStructure

MsgDescStructure

GetMsgOptsStructure


Reference Feedback


Last updated: Monday, September 23, 2013



MQCLOSE

Application has started the MQCLOSE MQI function

ObjectHandle

CloseOptions

CompCode

Reason

ResolvedQName

ResObjectString

ResolvedType


Reference Feedback


Last updated: Monday, September 23, 2013



MQCMIT

Application has started the MQCMIT MQI function

CompCode

Reason


Reference Feedback


Last updated: Monday, September 23, 2013



MQCONN and MQCONNX

Application has started the MQCONN or MQCONNX MQI function

ConnectionId

QueueManagerName

CompCode

Reason

ConnectOptions

ConnectionOptionsStructure

ChannelDefinitionStructure


Reference Feedback


Last updated: Monday, September 23, 2013



MQCTL

Application has started the MQCTL MQI function

CompCode

Reason

CtlOperation


Reference Feedback


Last updated: Monday, September 23, 2013



MQDISC

Application has started the MQDISC MQI function

CompCode

Reason


Reference Feedback


Last updated: Monday, September 23, 2013



MQGET

Application has started the MQGET MQI function

ObjectHandle

GetOptions

CompCode

Reason

MsgBuffer

MsgLength

HighResTime

BufferLength

ObjectName

ResolvedQName

ReportOptions

MsgType

Expiry

Format

Priority

Persistence

MsgId

CorrelId

ReplyToQueue

ReplyToQMgr

CodedCharSetId

Encoding

PutDate

PutTime

ResolvedQName

ResObjectString

ResolvedType

PolicyName

XmitqMsgId

XmitqCorrelId

XmitqPutTime

XmitqPutDate

XmitqRemoteQName

XmitqRemoteQMgr

MsgDescStructure

GetMsgOptsStructure


Reference Feedback


Last updated: Monday, September 23, 2013



MQINQ

Application has started the MQINQ MQI function

ObjectHandle

CompCode

Reason

SelectorCount

Selectors

ResolvedQName

ResObjectString

ResolvedType

IntAttrCount

IntAttrs

CharAttrs


Reference Feedback


Last updated: Monday, September 23, 2013



MQOPEN

Application has started the MQOPEN MQI function

ObjectType

ObjectName

ObjectQMgrName

ObjectHandle

CompCode

Reason

OpenOptions

AlternateUserId

RecsPresent

KnownDestCount

UnknownDestCount

InvalidDestCount

DynamicQName

ResolvedLocalQName 1 2

ResolvedLocalQMgrName 1 2

ResolvedQName 1 2

ResolvedQMgrName 1 2

AlternateSecurityId

ObjectString

SelectionString

ResObjectString

ResolvedType


Application Activity Distribution List PCF Group Header Structure

If the MQOPEN function opens a distribution list, then the MQOPEN parameters includes one AppActivityDistList PCF group for each of the queues in the distribution list up to the number of structures numbered in RecsPresent. The Ap-pActivityDistList PCF group combines information from the MQOR, and MQRR structures to identify the queue name, and indicate the result of the open operation on the queue. An AppActivityDistList group always starts with the following MQCFGR structure:

AppActivityDistList group MQCFGR structure

MQCFGR field Value Description
Type MQCFT_GROUP  
StrucLength Length in bytes of the MQCFGR structure  
Parameter MQGACF_APP_DIST_LIST Distribution list group parameter
ParameterCount 4 The number of parameter structures following the MQCFGR structure that are contained within this group.

ObjectName

ObjectQMgrName

CompCode

Reason


Reference Feedback


Last updated: Monday, September 23, 2013



MQPUT

Application has started the MQPUT MQI function.

ObjectHandle

PutOptions

CompCode

Reason

MsgBuffer

MsgLength

RecsPresent

KnownDestCount

UnknownDestCount

InvalidDestCount

HighResTime

ObjectName

ResolvedQName

ResolvedQMgrName

ResolvedLocalQName 1

ResolvedLocalQMgrName 1

ReportOptions

MsgType

Expiry

Format

Priority

Persistence

MsgId

CorrelId

ReplyToQueue

ReplyToQMgr

CodedCharSetId

Encoding

PutDate

PutTime

ResolvedQName

ResObjectString

ResolvedType

PolicyName

XmitqMsgId

XmitqCorrelId

XmitqPutTime

XmitqPutDate

XmitqRemoteQName

XmitqRemoteQMgr

PutMsgOptsStructure


MQPUT Application Activity Distribution List PCF Group Header Structure

If the MQPUT function is putting to a distribution list, then the MQPUT parameters include one AppActivityDistList PCF group. For each of the queues in the distribution list, see Application Activity Distribution List PCF Group Header Structure . The AppActivityDistList PCF group combines information from the MQPMR, and MQRR structures to identify the PUT parameters, and indicate the result of the PUT operation on each queue. For MQPUT operations the AppActivityDistList group contains some or all of the following parameters (the CompCode and Reason is present if the reason code is MQRC_MULTIPLE_REASONS and the other parameters are determined by the MQPMO.PutMsgRecFields field):

CompCode

Reason

MsgId

CorrelId

GroupId

Feedback

AccountingToken


Reference Feedback


Last updated: Monday, September 23, 2013



MQPUT1

Application has started the MQPUT1 MQI function

ObjectType

ObjectName

ObjectQMgrName

CompCode

Reason

PutOptions

AlternateUserId

RecsPresent

KnownDestCount

UnknownDestCount

InvalidDestCount

MsgBuffer

MsgLength

HighResTime

ResolvedQName

ResolvedQMgrName

ResolvedLocalQName 1

ResolvedLocalQMgrName 1

AlternateSecurityId

ObjectString

ResObjectString

ResolvedType

ReportOptions

MsgType

Expiry

Format

Priority

Persistence

MsgId

CorrelId

ReplyToQueue

ReplyToQMgr

CodedCharSetId

Encoding

PutDate

PutTime

PolicyName

XmitqMsgId

XmitqCorrelId

XmitqPutTime

XmitqPutDate

XmitqRemoteQName

XmitqRemoteQMgr

PutMsgOptsStructure


MQPUT1 AppActivityDistList PCF Group Header Structure

If the MQPUT1 function is putting to a distribution list, then the variable parameters include one AppActivityDistList PCF group. For each of the queues in the distribution list, see Application Activity Distribution List PCF Group Header Structure . The AppActivityDistList PCF group combines information from the MQOR, MQPMR, and MQRR structures to identify the objects, and the PUT parameters , and indicate the result of the PUT operation on each queue. For MQPUT1 operations the AppActivityDistList group contains some or all of the following parameters (the CompCode, Reason, ObjectName, and ObjectQMgrName is present if the reason code is MQRC_MULTIPLE_REASONS and the other parameters is determined by the MQPMO.PutMsgRecFields field):

CompCode

Reason

ObjectName

MsgId

CorrelId

GroupId

Feedback

AccountingToken


Reference Feedback


Last updated: Monday, September 23, 2013



MQSET

Application has started the MQSET MQI function

ObjectHandle

CompCode

Reason

SelectorCount

Selectors

ResolvedQName

ResObjectString

ResolvedType

IntAttrCount

IntAttrs

CharAttrs


Reference Feedback


Last updated: Monday, September 23, 2013



MQSUB

Application has started the MQSUB MQI function

CompCode

Reason

SubHandle

ObjectHandle

Options

ObjectName

ObjectString

AlternateUserId

AlternateSecurityId

SubName

SubUserData

SubCorrelId

SelectionString

ResolvedQName

ResObjectString

ResolvedType

SubDescriptorStructure


Reference Feedback


Last updated: Monday, September 23, 2013



MQSUBRQ

Application has started the MQSUBRQ MQI function

CompCode

Reason

SubHandle

SubOptions

Action

NumPubs


Reference Feedback


Last updated: Monday, September 23, 2013



MQSTAT

Application has started the MQSTAT MQI function

CompCode

Reason

Type

StatusStructure


Reference Feedback


Last updated: Monday, September 23, 2013



Variable Parameters for Application Activity XA Operations

XA operations are API calls that applications can make to enable MQ to participate in a transaction. The parameters for each operation are defined in the following section.

The trace level indicates the level of trace granularity that is required for the parameters to be included in the trace. The possible trace level values are:

  1. Low

    The parameter is included when low , medium or high activity tracing is configured for an application. This setting means that a parameter is always included in the AppActivityData group for the operation. This set of parameters is sufficient to trace the MQI calls an application makes, and to see if they are successful.

  2. Medium

    The parameter is only included in the AppActivityData group for the operation when medium or high activity tracing is configured for an application. This set of parameters adds information about the resources, for example, queue and topic names used by the application.

  3. High

    The parameter is only included in the AppActivityData group for the operation when high activity tracing is configured for an application. This set of parameters includes memory dumps of the structures passed to the MQI and XA functions. For this reason, it contains more information about the parameters used in MQI and XA calls. The structure memory dumps are shallow copies of the structures. To avoid erroneous attempts to dereference pointers, the pointer values in the structures are set to NULL.

    Note: The version of the structure that is dumped is not necessarily identical to the version used by an application. The structure can be modified by an API crossing exit, by the activity trace code, or by the queue manager. A queue manager can modify a structure to a later version, but the queue manager never changes it to an earlier version of the structure. To do so, would risk losing data.


AXREG

Application has started the AXREG AX function

XID

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



AXUNREG

Application has started the AXUNREG AX function

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XACLOSE

Application has started the XACLOSE AX function

Xa_info

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XACOMMIT

Application has started the XACOMMIT AX function

XID

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XACOMPLETE

Application has started the XACOMPLETE AX function

Handle

Retval

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XAEND

Application has started the XAEND AX function

XID

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XAFORGET

Application has started the AXREG AX function

XID

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XAOPEN

Application has started the XAOPEN AX function

Xa_info

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XAPREPARE

Application has started the XAPREPARE AX function

XID

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XARECOVER

Application has started the XARECOVER AX function

Count

XIDs

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XAROLLBACK

Application has started the XAROLLBACK AX function

XID

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



XASTART

Application has started the XASTART AX function

XID

Rmid

Flags

XARetCode


Reference Feedback


Last updated: Monday, September 23, 2013



Real-time monitoring

Real-time monitoring is a technique that allows you to determine the current state of queues and channels within a queue manager. The information returned is accurate at the moment the command was issued.

A number of commands are available that when issued return real-time information about queues and channels. Information can be returned for one or more queues or channels and can vary in quantity. Real-time monitoring can be used in the following tasks:

With real-time monitoring, information can be returned for either queues or channels. The amount of real-time information returned is controlled by queue manager, queue, and channel attributes.

Real-time monitoring for queues and channels is in addition to, and separate from, performance and channel event monitoring.


Attributes that control real-time monitoring

Some queue and channel status attributes hold monitoring information, if real-time monitoring is enabled. If real-time monitoring is not enabled, no monitoring information is held in these monitoring attributes. Examples demonstrate how you can use these queue and channel status attributes.

You can enable or disable real-time monitoring for individual queues or channels, or for multiple queues or channels. To control individual queues or channels, set the queue attribute MONQ or the channel attribute MONCH, to enable or disable real-time monitoring. To control many queues or channels together, enable or disable real-time monitoring at the queue manager level by using the queue manager attributes MONQ and MONCHL. For all queue and channel objects with a monitoring attribute that is specified with the default value, QMGR, real-time monitoring is controlled at the queue manager level.

Automatically defined cluster-sender channels are not WebSphere MQ objects, so do not have attributes in the same way as channel objects. To control automatically defined cluster-sender channels, use the queue manager attribute, MONACLS. This attribute determines whether automatically defined cluster-sender channels within a queue manager are enabled or disabled for channel monitoring.

For real-time monitoring of channels, you can set the MONCHL attribute to one of the three monitoring levels: low, medium, or high. You can set the monitoring level either at the object level or at the queue manager level. The choice of level is dependent on your system. Collecting monitoring data might require some instructions that are relatively expensive computationally, such as obtaining system time. To reduce the effect of real-time monitoring, the medium and low monitoring options measure a sample of the data at regular intervals rather than collecting data all the time. Table 1 summarizes the monitoring levels available for real-time monitoring of channels:

Monitoring levels

Level Description Usage
Low

Measure a small sample of the data, at regular intervals.

For objects that process a high volume of messages.

Medium

Measure a sample of the data, at regular intervals.

For most objects.

High

Measure all data, at regular intervals.

For objects that process only a few messages per second, on which the most current information is important.

For real-time monitoring of queues, you can set the MONQ attribute to one of the three monitoring levels, low, medium or high. However, there is no distinction between these values. The values all enable data collection, but do not affect the size of the sample.


Examples

The following examples demonstrate how to set the necessary queue, channel, and queue manager attributes to control the level of monitoring. For all of the examples, when monitoring is enabled, queue and channel objects have a medium level of monitoring.

  1. To enable both queue and channel monitoring for all queues and channels at the queue manager level, use the following commands:

     ALTER QMGR MONQ(MEDIUM) MONCHL(MEDIUM)
    ALTER QL(Q1) MONQ(QMGR)
    ALTER CHL(QM1.TO.QM2) CHLTYPE(SDR) MONCHL(QMGR)
    
    
  2. To enable monitoring for all queues and channels, with the exception of local queue, Q1, and sender channel, QM1.TO.QM2, use the following commands:

     ALTER QMGR MONQ(MEDIUM) MONCHL(MEDIUM)
    ALTER QL(Q1) MONQ(OFF)
    ALTER CHL(QM1.TO.QM2) CHLTYPE(SDR) MONCHL(OFF)
    
  3. To disable both queue and channel monitoring for all queues and channels, with the exception of local queue, Q1, and sender channel, QM1.TO.QM2, use the following commands:

     ALTER QMGR MONQ(OFF) MONCHL(OFF)
    ALTER QL(Q1) MONQ(MEDIUM)
    ALTER CHL(QM1.TO.QM2) CHLTYPE(SDR) MONCHL(MEDIUM)
    
  4. To disable both queue and channel monitoring for all queues and channels, regardless of individual object attributes, use the following command:

     ALTER QMGR MONQ(NONE) MONCHL(NONE)
    
  5. To control the monitoring capabilities of automatically defined cluster-sender channels Use the following command:

     ALTER QMGR MONACLS(MEDIUM)
    
  6. To specify that automatically defined cluster-sender channels are to use the queue manager setting for channel monitoring, use the following command:

     ALTER QMGR MONACLS(QMGR)
    


Display queue and channel monitoring data

To display real-time monitoring information for a queue or channel, use either the WebSphere MQ Explorer or the appropriate MQSC command. Some monitoring fields display a comma-separated pair of indicator values, which help you to monitor the operation of your queue manager. Examples demonstrate how you can display monitoring data. Monitoring fields that display a pair of values separated by a comma provide short term and long term indicators for the time measured since monitoring was enabled for the object, or from when the queue manager was started:

These indicator values are most useful to detect changes in the operation of your queue manager. This requires knowledge of the times these indicators show when in normal use, in order to detect increases in these times. By collecting and checking these values regularly you can detect fluctuations in the operation of your queue manager. This can indicate a change in performance.

Obtain real-time monitoring information as follows:


Procedure

  1. To display real-time monitoring information for a queue, use either the WebSphere MQ Explorer or the MQSC command DISPLAY QSTATUS, specifying the optional parameter MONITOR.
  2. To display real-time monitoring information for a channel, use either the WebSphere MQ Explorer or the MQSC command DISPLAY CHSTATUS, specifying the optional parameter MONITOR.


Example

The queue, Q1, has the attribute MONQ set to the default value, QMGR, and the queue manager that owns the queue has the attribute MONQ set to MEDIUM. To display the monitoring fields collected for this queue, use the following command:

 DISPLAY QSTATUS(Q1) MONITOR
The monitoring fields and monitoring level of queue, Q1 are displayed as follows:

 QSTATUS(Q1)
TYPE(QUEUE)
MONQ(MEDIUM)
QTIME(11892157,24052785)
MSGAGE(37)
LPUTDATE(2005-03-02)
LPUTTIME(09.52.13)
LGETDATE(2005-03-02)
LGETTIME(09.51.02)

The sender channel, QM1.TO.QM2, has the attribute MONCHL set to the default value, QMGR, and the queue manager that owns the queue has the attribute MONCHL set to MEDIUM. To display the monitoring fields collected for this sender channel, use the following command:

 DISPLAY CHSTATUS(QM1.TO.QM2) MONITOR
The monitoring fields and monitoring level of sender channel, QM1.TO.QM2 are displayed as follows:

 CHSTATUS(QM1.TO.QM2)
XMITQ(Q1)
CONNAME(127.0.0.1)
CURRENT
CHLTYPE(SDR)
STATUS(RUNNING)
SUBSTATE(MQGET)
MONCHL(MEDIUM)
XQTIME(755394737,755199260)
NETTIME(13372,13372)
EXITTIME(0,0)
XBATCHSZ(50,50)
COMPTIME(0,0)
STOPREQ(NO)
RQMNAME(QM2) 


Monitoring queues

Use this page to view tasks that help you to resolve a problem with a queue and the application that services that queue. Various monitoring options are available to determine the problem

Frequently, the first sign of a problem with a queue that is being serviced is that the number of messages on the queue (CURDEPTH) increases. If you expect an increase at certain times of day or under certain workloads, an increasing number of messages might not indicate a problem. However, if you have no explanation for the increasing number of messages, you might want to investigate the cause.

You might have an application queue where there is a problem with the application, or a transmission queue where there is a problem with the channel. Additional monitoring options are available when the application that services the queue is a channel.

The following examples investigate problems with a particular queue, called Q1, and describe the fields that you look at in the output of various commands:


Determining whether your application has the queue open

If you have a problem with a queue, check whether your application has the queue open Perform the following steps to determine whether your application has the queue open:


Procedure

  1. Ensure that the application that is running against the queue is the application that you expect. Issue the following command for the queue in question:

     DISPLAY QSTATUS(Q1) TYPE(HANDLE) ALL
    
    In the output, look at the APPLTAG field, and check that the name of your application is shown. If the name of your application is not shown, or if there is no output at all, start your application.

  2. If the queue is a transmission queue, look in the output at the CHANNEL field. If the channel name is not shown in the CHANNEL field, determine whether the channel is running.

  3. Ensure that the application that is running against the queue has the queue open for input. Issue the following command:

     DISPLAY QSTATUS(Q1) TYPE(QUEUE) ALL
    
    In the output, look at the IPPROCS field to see if any application has the queue open for input. If the value is 0 and this is a user application queue, make sure that the application opens the queue for input to get the messages off the queue.


Task Feedback


Last updated: Monday, September 23, 2013



Checking that messages on the queue are available

If you have a large number of messages on the queue and your application is not processing any of those messages, check whether the messages on the queue are available to your application Perform the following steps to investigate why your application is not processing messages from the queue:


Procedure

  1. Ensure that your application is not asking for a specific message ID or correlation ID when it should be processing all the messages on the queue.

  2. Although the current depth of the queue might show that there is an increasing number of messages on the queue, some messages on the queue might not be available to be got by an application, because they are not committed; the current depth includes the number of uncommitted MQPUTs of messages to the queue. Issue the following command:

     DISPLAY QSTATUS(Q1) TYPE(QUEUE) ALL
    
    In the output, look at the UNCOM field to see whether there are any uncommitted messages on the queue.

  3. If your application is attempting to get any messages from the queue, check whether the putting application is committing the messages correctly. Issue the following command to find out the names of applications that are putting messages to this queue:

     DISPLAY QSTATUS(Q1) TYPE(HANDLE) OPENTYPE(OUTPUT)
    

  4. Then issue the following command, inserting in <appltag> the APPLTAG value from the output of the previous command:

     DISPLAY CONN(*) WHERE(APPLTAG EQ <appltag>) UOWSTDA UOWSTTI
    
    This shows when the unit of work was started and will help you discover whether the application is creating a long running unit of work. If the putting application is a channel, you might want to investigate why a batch is taking a long time to complete.


Task Feedback


Last updated: Monday, September 23, 2013



Checking whether your application is getting messages off the queue

If you have a problem with a queue and the application that services that queue, check whether your application is getting messages off the queue To check whether your application is getting messages off the queue, perform the following checks:


Procedure

  1. Ensure that the application that is running against the queue is actually processing messages from the queue. Issue the following command:

     DISPLAY QSTATUS(Q1) TYPE(QUEUE) ALL
    
    In the output, look at the LGETDATE and LGETTIME fields which show when the last get was done from the queue.

  2. If the last get from this queue was longer ago than expected, ensure that the application is processing messages correctly. If the application is a channel, check whether messages are moving through that channel


Task Feedback


Last updated: Monday, September 23, 2013



Determining whether the application can process messages fast enough

If messages are building up on the queue, but your other checks have not found any processing problems, check that the application can process messages fast enough. If the application is a channel, check that the channel can process messages fast enough. To determine whether the application is processing messages fast enough, perform the following tests:


Procedure

  1. Issue the following command periodically to gather performance data about the queue:

     DISPLAY QSTATUS(Q1) TYPE(QUEUE) ALL
    
    If the values in the QTIME indicators are high, or are increasing over the period, and you have already ruled out the possibility of long running Units of Work by checking that messages on the queue are available, the getting application might not be keeping up with the putting applications.

  2. If your getting application cannot keep up with the putting applications, consider adding another getting application to process the queue. Whether you can add another getting application depends on the design of the application and whether the queue can be shared by more than one application. Features such as message grouping or getting by correlation ID might help to ensure that two applications can process a queue simultaneously.


Task Feedback


Last updated: Monday, September 23, 2013



Checking the queue when the current depth is not increasing

Even if the current depth of your queue is not increasing, it might still be useful to monitor the queue to check whether your application is processing messages correctly. To gather performance data about the queue: Issue the following command periodically:


Procedure

Issue the following command periodically:

 DISPLAY QSTATUS(Q1) TYPE(QUEUE) MSGAGE QTIME
In the output, if the value in MSGAGE increases over the period of time, and your application is designed to process all messages, this might indicate that some messages are not being processed at all.


Task Feedback


Last updated: Monday, September 23, 2013



Monitoring channels

Use this page to view tasks that help you to resolve a problem with a transmission queue and the channel that services that queue. Various channel monitoring options are available to determine the problem.

Frequently, the first sign of a problem with a queue that is being serviced is that the number of messages on the queue (CURDEPTH) increases. If you expect an increase at certain times of day or under certain workloads, an increasing number of messages might not indicate a problem. However, if you have no explanation for the increasing number of messages, you might want to investigate the cause.

You might have a problem with the channel that services a transmission queue. Various channel monitoring options are available to help you to determine the problem.

The following examples investigate problems with a transmission queue called QM2 and a channel called QM1.TO.QM2. This channel is used to send messages from queue manager, QM1, to queue manager, QM2. The channel definition at queue manager QM1 is either a sender or server channel, and the channel definition at queue manager, QM2, is either a receiver or requester channel.


Determining whether the channel is running

If you have a problem with a transmission queue, check whether the channel is running. Perform the following steps to check the status of the channel that is servicing the transmission queue:


Procedure

  1. Issue the following command to find out which channel you expect to process the transmission queue QM2:

     DIS CHANNEL(*) WHERE(XMITQ EQ QM2)
    
    In this example, the output of this command shows that the channel servicing the transmission queue is QM1.TO.QM2

  2. Issue the following command to determine the status of the channel, QM1.TO.QM2:

     DIS CHSTATUS(QM1.TO.QM2) ALL
    

  3. Inspect the STATUS field of the output from the CHSTATUS command:
    • If the value of the STATUS field is RUNNING, check that the channel is moving messages
    • If the output from the command shows no status, or the value of the STATUS field is STOPPED, RETRY, BINDING, or REQUESTING, perform the appropriate step, as follows:

  4. Optional: If the value of the STATUS field shows no status, the channel is inactive, so perform the following steps:

    1. If the channel should have been started automatically by a trigger, check that the messages on the transmission queue are available. If there are messages available on the transmission queue, check that the trigger settings on the transmission queue are correct.

    2. Issue the following command to start the channel again manually:

       START CHANNEL(QM1.TO.QM2)
      

  5. Optional: If the value of the STATUS field is STOPPED, perform the following steps:

    1. Check the error logs to determine why the channel stopped. If the channel stopped owing to an error, correct the problem. Ensure also that the channel has values specified for the retry attributes: SHORTRTY and LONGRTY. In the event of transient failures such as network errors, the channel will then attempt to restart automatically.

    2. Issue the following command to start the channel again manually:

       START CHANNEL(QM1.TO.QM2)
      

  6. Optional: If the value of the STATUS field is RETRY, perform the following steps:

    1. Check the error logs to identify the error, then correct the problem.

    2. Issue the following command to start the channel again manually:

       START CHANNEL(QM1.TO.QM2)
      
      or wait for the channel to connect successfully on its next retry.

  7. Optional: If the value of the STATUS field is BINDING or REQUESTING, the channel has not yet successfully connected to the partner. Perform the following steps:

    1. Issue the following command, at both ends of the channel, to determine the substate of the channel:

       DIS CHSTATUS(QM1.TO.QM2) ALL
      

      Note:

      1. In some cases there might be a substate at one end of the channel only.
      2. Many substates are transitory, so issue the command a few times to detect whether a channel is stuck in a particular substate.

    2. Check Table 1 to determine what action to take:

      Substates seen with status binding or requesting

      Initiating MCA substate 1 Responding MCA substate 2 Notes
      NAMESERVER   The initiating MCA is waiting for a name server request to complete. Ensure that the correct host name has been specified in the channel attribute, CONNAME, and that your name servers are set up correctly.
      SCYEXIT SCYEXIT The MCAs are currently in conversation through a security exit. For more information, see Determining whether the channel can process messages fast enough .
        CHADEXIT The channel autodefinition exit is currently executing. For more information, see Determining whether the channel can process messages fast enough .

      RCVEXIT
      SENDEXIT
      MSGEXIT
      MREXIT

      RCVEXIT
      SENDEXIT
      MSGEXIT
      MREXIT

      Exits are called at channel startup for MQXR_INIT. Review the processing in this part of your exit if this takes a long time. For more information, see Determining whether the channel can process messages fast enough .
      SERIALIZE SERIALIZE This substate only applies to channels with a disposition of SHARED.
      NETCONNECT   This substate is shown if there is a delay in connecting due to incorrect network configuration.
      SSLHANDSHAKE SSLHANDSHAKE An SSL handshake consists of a number of sends and receives. If network times are slow, or connection to lookup CRLs are slow, this affects the time taken to do the handshake.

      Notes:

      1. The initiating MCA is the end of the channel which started the conversation. This can be senders, cluster-senders, fully-qualified servers and requesters. In a server-requester pair, it is the end from which you started the channel.
      2. The responding MCA is the end of the channel which responded to the request to start the conversation. This can be receivers, cluster-receivers, requesters (when the server or sender is started), servers (when the requester is started) and senders (in a requester-sender call-back pair of channels).


Task Feedback


Last updated: Monday, September 23, 2013



Checking that the channel is moving messages

If you have a problem with a transmission queue, check that the channel is moving messages Issue the command DIS CHSTATUS(QM1.TO.QM2) ALL. If the value of the STATUS field is RUNNING, the channel has successfully connected to the partner system.

Check that there are no uncommitted messages on the transmission queue, as described in Checking that messages on the queue are available . If there are messages available for the channel to get and send, perform the following checks:


Procedure

  1. In the output from the display channel status command, DIS CHSTATUS(QM1.TO.QM2) ALL, look at the following fields:

    MSGS

    Number of messages sent or received (or, for server-connection channels, the number of MQI calls handled) during this session (since the channel was started).

    BUFSSENT

    Number of transmission buffers sent. This includes transmissions to send control information only.

    BYTSSENT

    Number of bytes sent during this session (since the channel was started). This includes control information sent by the message channel agent.

    LSTMSGDA

    Date when the last message was sent or MQI call was handled, see LSTMSGTI.

    LSTMSGTI

    Time when the last message was sent or MQI call was handled. For a sender or server, this is the time the last message (the last part of it if it was split) was sent. For a requester or receiver, it is the time the last message was put to its target queue. For a server-connection channel, it is the time when the last MQI call completed.

    CURMSGS

    For a sending channel, this is the number of messages that have been sent in the current batch. For a receiving channel, it is the number of messages that have been received in the current batch. The value is reset to zero, for both sending and receiving channels, when the batch is committed.

  2. Determine whether the channel has sent any messages since it started. If any have been sent, determine when the last message was sent.

  3. If the channel has started a batch that has not yet completed, as indicated by a non-zero value in CURMSGS, the channel might be waiting for the other end of the channel to acknowledge the batch. Look at the SUBSTATE field in the output and refer to Table 1 :

    Sender and receiver MCA substates

    Sender SUBSTATE Receiver SUBSTATE Notes
    MQGET RECEIVE Normal states of a channel at rest.
    SEND RECEIVE SEND is usually a transitory state. If SEND is seen it indicates that the communication protocol buffers have filled. This can indicate a network problem.
    RECEIVE   If the sender is seen in RECEIVE substate for any length of time, it is waiting on a response, either to a batch completion or a heartbeat. You might want to check why a batch takes a long time to complete.

    Note: You might also want to determine whether the channel can process messages fast enough, especially if the channel has a substate associated with exit processing.


Task Feedback


Last updated: Monday, September 23, 2013



Checking why a batch takes a long time to complete

Use this page to view some reasons why a batch can take a long time to complete. When a sender channel has sent a batch of messages it waits for confirmation of that batch from the receiver, unless the channel is pipelined. The following factors can affect how long the sender channel waits:


Procedure


Task Feedback


Last updated: Monday, September 23, 2013



Determining whether the channel can process messages fast enough

If there messages are building up on the transmission queue, but you have found no processing problems, determine whether the channel can process messages fast enough. Issue the following command repeatedly over a period of time to gather performance data about the channel:

 DIS CHSTATUS(QM1.TO.QM2) ALL
Confirm that there are no uncommitted messages on the transmission queue, as described in Checking that messages on the queue are available , then check the XQTIME field in the output from the display channel status command. When the values of the XQTIME indicators are consistently high, or increase over the measurement period, the indication is that the channel is not keeping pace with the putting applications.

Perform the following tests:


Procedure

  1. Check whether exits are processing. If exits are used on the channel that is delivering these messages, they might add to the time spent processing messages. To identify if this is the case, do the following checks:

    1. In the output of the command DIS CHSTATUS(QM1.TO.QM2) ALL, check the EXITTIME field. If the time spent in exits is higher than expected, review the processing in your exits for any unnecessary loops or extra processing, especially in message, send, and receive exits. Such processing affects all messages moved across the channel.

    2. In the output of the command DIS CHSTATUS(QM1.TO.QM2) ALL, check the SUBSTATE field. If the channel has of one of the following substates for a significant time, review the processing in your exits:

  2. Check whether the network is slow. If messages are not moving fast enough across a channel, it might be because the network is slow. To identify if this is the case, do the following checks:

    1. In the output of the command DIS CHSTATUS(QM1.TO.QM2) ALL, check the NETTIME field. These indicators are measured when the sending channel asks its partner for a response. This happens at the end of each batch and, when a channel is idle during heartbeating.
    2. If this indicator shows that round trips are taking longer than expected, use other network monitoring tools to investigate the performance of your network.

  3. Check whether the channel is using compression. If the channel is using compression, this adds to the time spent processing messages. If the channel is using only one compression algorithm, do the following checks:

    1. In the output of the command DIS CHSTATUS(QM1.TO.QM2) ALL, check the COMPTIME field. These indicators show the time spent during compression or decompression.
    2. If the chosen compression is not reducing the amount of data to send by the expected amount, change the compression algorithm.

  4. If the channel is using multiple compression algorithms, do the following checks:

    1. In the output of the command DIS CHSTATUS(QM1.TO.QM2) ALL, check the COMPTIME, COMPHDR, and COMPMSG fields.
    2. Change the compression algorithms specified on the channel definition, or consider writing a message exit to override the channel's choice of compression algorithm for particular messages if the rate of compression, or choice of algorithm, is not providing the required compression or performance.


Task Feedback


Last updated: Monday, September 23, 2013



Solving problems with cluster channels

If you have a build up of messages on the SYSTEM.CLUSTER.TRANSMIT.QUEUE queue, the first step in diagnosing the problem is discovering which channel, or channels, are having a problem delivering messages. To discover which channel, or channels, using the SYSTEM.CLUSTER.TRANSMIT.QUEUE are having a problem delivering messages. Perform the following checks:


Procedure

  1. Issue the following command:

     DIS CHSTATUS(*) WHERE(XQMSGSA GT 1)
    

    Note: If you have a busy cluster that has many messages moving, consider issuing this command with a higher number to eliminate the channels that have only a few messages available to deliver.

  2. Look through the output for the channel, or channels, that have large values in the field XQMSGSA. Determine why the channel is not moving messages, or is not moving them fast enough. Use the tasks outlined in Monitoring channels to diagnose the problems with the channels found to be causing the build up.


Task Feedback


Last updated: Monday, September 23, 2013



The Windows performance monitor

In WebSphere MQ Version 7.0 and earlier versions, it was possible to monitor the performance of local queues on Windows systems by using the Windows performance monitor. As of WebSphere MQ Version 7.1, this method of performance monitoring is no longer available.

You can monitor queues on all supported platforms by using methods described in Real-time monitoring .