Use automatic cluster configuration

You configure IBM MQ to enable automatic configuration by changing the qm.ini configuration information.

Note: We can only use the AutoCluster stanza for uniform clusters.

Stanzas to configure

We can change the following stanzas:

AutoConfig: Defined in the qm.ini file. When the queue manager starts, it identifies which automatic configuration files to apply.; We should use this mechanism to distribute identical cluster configuration when uniform clusters are being used.
AutoCluster: Defined in the qm.ini file. Used when the queue manager starts to identify if the cluster is a member of an automatic cluster, and can identify the full repositories of the cluster.
Variables: Defined in the qm.ini file. Contains some queue manager variables.

Attributes for the AutoConfig stanza

The following two attributes are permitted in the AutoConfig stanza:

MQSCConfig=<Path>: The path is either a full file path, or path to a directory, where all files *.mqsc are applied to the queue manager, on every queue manager start.
For more information, see Automatic configuration from an MQSC script at startup.
IniConfig=<Path>: The path is either a full file path, or path to a directory, where all files *.ini are applied to the qm.ini file, on every queue manager start.
For more information, see Automatic configuration of qm.ini at startup.

These attributes are frequently used as part of setting up uniform clusters. For more information, see Create a new uniform cluster.

Example stanza:

AutoConfig:
MQSCConfig=C:\MQ_Configuration\uniclus.mqsc   
IniConfig=C:\MQ_Configuration\uniclus.ini

Attributes for the AutoCluster stanza

The following attributes are mandatory for the AutoCluster stanza:

Type=Uniform: Specifies the type of automatic cluster, and the only valid option is Uniform, which represents a uniform cluster.
ClusterName=<String>: Name of the cluster, that is the automatic cluster name.

The presence of the above attributes enables application balancing for uniform clusters. See Automatic application balancing for further details.

In addition, simplified setup of a cluster can be performed if the cluster is described in this stanza. For more information, see Create a new uniform cluster. When this is used, we can name two queue managers, and give their connection names for the full repositories for this automatic cluster.

The following attributes are optional for the AutoCluster stanza but we must provide them in pairs:

RepositoryName1=<String>: This is the queue manager name for the first full repository in the automatic cluster. This can be the name of this queue manager, or another one.
Repository1Conname=<Connection Name String>: This is the connection name (CONNAME) value for how members of the automatic cluster should connect to this queue manager.

Additionally, we can identify a second full repository for the cluster:

Repository2Name=<String>
Repository2Conname=<Connection name string>

Example stanza:

AutoCluster:
   Repository2Conname=myFR1.hostname(1414)
   Repository2Name=QMFR1
   Repository1Conname= myFR2.hostname(1414)
   Repository1Name=QMFR2
   ClusterName=UNICLUS
   Type=Uniform

Attributes for the Variables stanza

An attribute=value pair is valid in the attribute field. These can be supplied using the -iv command line option on the crtmqm command when you create a queue manager.

We can use attributes listed in the Variables stanza during automatic cluster configuration of CONNAME and the channel name MQSC fields of a cluster receiver channel.

Parent topic: Configure a uniform cluster