Recovery log manager messages (CSQJ...)
CSQJ001ICURRENT COPY n ACTIVE LOG DATA SET IS DSNAME=dsname, STARTRBA=sss ENDRBA=tttExplanation:This message is generated for one of two reasons:
- When the queue manager starts, this information message is sent to identify the current active log data sets (copy 1 and, if dual logging is used, copy 2).
- When the current active log data set is full (or when an ARCHIVE LOG command is issued), MQ will switch to the next available active log data set. This message identifies the next available active log data set that will be used for logging.
The value specified by STARTRBA is the RBA of the first byte of log data in the named data set. The value specified by ENDRBA is the RBA of the last possible byte in the data set.
System Programmer Response:None required. However, if recovery is required, information from this message might be required as input to the change log inventory utility (CSQJU003).
CSQJ002IEND OF ACTIVE LOG DATA SET DSNAME=dsname, STARTRBA=sss ENDRBA=tttExplanation:This message is sent when logging switches to a new empty data set. The message shows the name and log RBA range of the full data set.
System Programmer Response:None required. However, if recovery is required, information from this message might be required as input to the change log inventory utility (CSQJU003).
CSQJ003IFULL ARCHIVE LOG VOLUME DSNAME=dsname, STARTRBA=sss ENDRBA=ttt, STARTTIME=ppp ENDTIME=qqq, UNIT=unitname, COPYnVOL=vvv VOLSPAN=xxx CATLG=yyyExplanation:Offloading for the specified archive log data set was successfully completed for the given volume. If the data set spans multiple tape volumes, this message is generated for each tape volume.
System Action:An archive log data set has been created, and the archive log data set inventory in the BSDS has been updated with the information in the message:
- DSNAME
- The name of the archive log data set
- STARTRBA
- The starting RBA contained in the volume
- ENDRBA
- The ending RBA contained in the volume
- STARTTIME
- The starting store-clock value of the log records in the volume
- ENDTIME
- The ending store-clock value of the log records in the volume
- UNIT
- The device unit to which the data set was allocated
- COPYnVOL
- The name of the volume; this is displayed as COPY1VOL if this is the copy-1 archive log data set, and as COPY2VOL if this is the copy-2 archive log data set
- VOLSPAN
- An indicator to denote one of four conditions:
- NO
- The data set is entirely contained on the volume specified by COPYnVOL
- FIRST
- This is the first entry of a multivolume data set
- MIDDLE
- This is the middle entry of a multivolume data set
- LAST
- This is the last entry of a multivolume data set
- CATLG
- An indicator to denote one of two conditions:
- NO
- The archive log data set is uncataloged
- YES
- The archive log data set is cataloged
The BSDS is automatically updated with the information contained in this message; however, if recovery is required, information from this message might be required as input to the change log inventory utility (CSQJU003).
CSQJ004IACTIVE LOG COPY n INACTIVE, LOG IN SINGLE MODE, ENDRBA=tttExplanation:This message is sent when the dual active logging option is selected and copy n becomes inactive. A log copy becomes inactive when the next active log data set is not ready when required. ENDRBA is the last byte of log data written on copy n. This is usually caused by a delay in offload.
System Action:The log is switched to single mode until the next data set for copy n is ready for logging.
If the queue manager is shut down or terminates abnormally while in single mode with the system parameter option still set for dual active data sets, the previous state of the active log data sets determines what happens when the queue manager is started, as follows:
- If fewer than two data sets are available (not flagged as STOPPED) for each set of active logs, queue manager startup terminates and message CSQJ112E is issued.
- If an active log data set is in NOTREUSABLE state, the queue manager can be started in single logging mode, but dual mode takes effect when the other active log data set becomes available after offloading.
System Programmer Response:Perform a display request to ensure that there are no outstanding requests that are related to the log offload process. Take the necessary action to satisfy any requests, and permit offload to continue.
If the switch to single mode was caused by the lack of a resource required for offload, the necessary resource should be made available to allow offload to complete and thus permit dual logging to proceed. If recovery is required, information from this message might be required as input to the change log inventory utility (CSQJU003).
CSQJ005IACTIVE LOG COPY n IS ACTIVE, LOG IN DUAL MODE, STARTRBA=sssExplanation:This message is sent when copy n of the log becomes active after previously being flagged as inactive. STARTRBA is the RBA of the first byte of log data written on copy n after it was activated.
System Programmer Response:None required. However, if recovery is required, information from this message might be required as input to the change log inventory utility (CSQJU003).
CSQJ006IALLOCATION FOR NEW ARCHIVE LOG DATA SET HAS BEEN CANCELED BY OPERATORExplanation:This message is sent if the operator answers 'N' to message CSQJ008E.
System Action:If the allocation is for the first copy of the archive log data set, offload terminates processing until the next time it is activated. If the first copy has already been allocated and this request is for the second copy, offload switches to single offload mode for this data set only.
CSQJ007IALLOCATION FOR ARCHIVE VOL SER=volser HAS BEEN CANCELED BY OPERATORExplanation:If the operator answers 'N' to message CSQJ009E, this message is issued. volser is the volume serial of an archive log volume required to satisfy the read request. The name of the archive data set is given by message CSQJ022I which follows.
System Action:The read request that needed the archive volume is unsuccessful. If the request was issued with the COND=YES parameter, the log manager returns to its invoker with return code 12 and reason code X'00D1032B'. Otherwise, the log manager's invoker ends abnormally with the same reason code.
CSQJ008Enn OF mm ACTIVE LOGS ARE FULL, qmgr-name NEEDS ARCHIVE SCRATCHExplanation:MQ needs a scratch volume for offloading an active log data set. qmgr-name is the name of the queue manager. nn is the number of full active log data sets. mm is the total number of active log data sets.
System Action:The offload task issues message CSQJ021D and waits for the operator's reply.
Operator Response:There are three options:
- Get a scratch volume ready, make sure there is an available unit for the volume, and reply 'Y'. MQ then continues with the offload.
- Determine from the number of active log data sets available whether the offload can be delayed until the next time an active log data set becomes full. If the process can be delayed, reply 'N'.
This response has two possible effects:
- If dual archive logging is in effect and this allocation is for a copy 1 archive data set, the 'N' response delays the offload process until the next active log data set becomes full.
- However, if the copy 1 archive data set has already been allocated and this request is for copy 2, the 'N' response causes the offload to switch to single archive mode (the switch is for this data set only).
- Defer giving a response. This causes offload to wait before processing. However, because offload is a separate service task, the wait does not affect MQ performance.
If offloading to DASD, an error has occurred attempting to allocate an archive log data set. Reply 'Y' to receive the error messages.
CSQJ009Eqmgr-name NEEDS VOL SER=nnnnnnExplanation:MQ needs the specified archive volume for a read operation. qmgr-name is the name of the queue manager.
System Action:The archive log read service task issues message CSQJ021D and waits for the operator's reply. This wait affects the agent for which the log read was issued and any other agents that might be waiting on the log read service task queue.
Operator Response:Locate the requested volume, ensure that a device is available, and reply 'Y'. MQ continues with dynamic allocation and begins reading the log.
If dual archiving is in effect, a response of 'N' causes archive read to reissue the message for the copy 2 archive VOLSER with the same RBA range. A response of 'N' to this second message, or to the initial message for single archiving, causes the archive read service task to be unsuccessful, with unpredictable results.
CSQJ010IINVALID RESPONSE - NOT Y OR NExplanation:During archive data set allocation, a reply message was issued. The user did not respond correctly to the reply message. Either 'Y' or 'N' must be entered.
System Action:The original message is repeated.
Operator Response:Reply as indicated in the repeated message.
CSQJ011DRESTART CONTROL rrr CREATED AT date time FOUND. REPLY Y TO USE, N TO CANCELExplanation:During queue manager initialization, a conditional restart control record was found in the BSDS data set. Both the record identifier (a 4-byte hexadecimal number) and the creation time stamp are displayed to help identify the conditional restart record which will be used. If you want a conditional restart using that record, reply 'Y' to the message. Otherwise, reply 'N'.
System Action:If 'Y' is the response, the queue manager is started conditionally, using the record found. If 'N' is the response, startup is terminated.
System Programmer Response:Respond as indicated.
If a normal restart has failed and you have created a conditional restart record with the change log inventory utility, check whether the time and date in the message agree with when you created that record. If they do, reply 'Y'. If they do not, reply 'N' and investigate the discrepancy.
CSQJ012EERROR ccc READING RBA rrr IN DATA SET dsname, CONNECTION-ID=xxxx THREAD-XREF=yyyyyyExplanation:While scanning log records read into a buffer, MQ detected a logical error with reason code ccc. rrr is the log RBA of the segment in the buffer at which the error was detected. dsname is the name of the active or archive log data set from which the record was read. If dsname is blank, the data was read from an active log output buffer.
The connection ID and thread-xref identify the user or application that encountered the problem. Messages that have the same connection ID and thread-xref relate to the same user.
System Action:The application program is terminated with reason code ccc. However, information in this message might be useful in diagnosing the abnormal termination that will follow.
System Programmer Response:See the WebSphere MQ for z/OS System Administration Guide for information about dealing with problems on the log.
CSQJ013ETERMINAL ERROR ccc IN BUFFER rrr BEFORE ACTIVE LOG WRITEExplanation:A scan of the log output buffer, just prior to writing the buffer, detected an inconsistency in the log data. ccc is the reason code associated with the SDUMP that is produced. rrr is the log RBA at which the error was detected.
System Action:The queue manager will terminate with a dump, and will not write the damaged buffer to either COPY 1 or COPY 2 active log data set.
System Programmer Response:Restart the queue manager after it terminates.
Because the damaged buffer has not been written to a log data set, the queue manager can be restarted. No corrective action is required.
CSQJ014ETERMINAL ERROR ccc IN BUFFER rrr AFTER ACTIVE LOG WRITEExplanation:A scan of the log output buffer, after writing to the first copy of the active log data set and before writing to the second copy, detected an inconsistency in the log data. ccc is the reason code associated with the SDUMP that is produced. rrr is the log RBA at which the error was detected.
System Action:The queue manager terminates with a dump, and does not write the damaged buffer to the COPY 2 data set.
System Programmer Response:The block containing the indicated log RBA might be damaged. The buffer was found to be in error at the completion of the write to the COPY 1 data set of the active log.
If dual active logs are being used, use the print log map utility (CSQJU004) to list the active log data sets for both copies of the active log. Find the COPY 2 data set with the corresponding RBA, and copy that data set (using Access Method Services REPRO) to the COPY 1 data set. Start the queue manager.
If only a single active log is used, contact the IBM support center for assistance. An attempt to start the queue manager might succeed if the damage to the buffer occurred after completion of the write to DASD.
CSQJ020Icsect-name RECEIVED REPLY OF N TO msg-num. QUEUE MANAGER STARTUP IS TERMINATEDExplanation:The operator chose to terminate queue manager startup by answering 'N' to msg-num.
System Action:The queue manager will not restart.
Operator Response:To restart the queue manager, follow the operator response given for message msg-num.
CSQJ021DREPLY Y WHEN DEVICE READY OR N TO CANCELExplanation:An archive log data set needs allocating, as indicated in the preceding CSQJ008E or CSQJ009E message.
System Action:The log service task waits for the operator's reply.
Operator Response:Refer to the explanation of message CSQJ008E or CSQJ009E as appropriate. When the device and volume is ready, reply 'Y'; otherwise, reply 'N' to cancel the operation.
CSQJ022IDSNAME=dsnameExplanation:dsname is the name of the archive data set to which the preceding message refers.
CSQJ030ERBA RANGE startrba TO endrba NOT AVAILABLE IN ACTIVE LOG DATA SETSExplanation:Previous errors have made the active log data sets (that contain the RBA range reported in the message) unavailable. The status of these logs is STOPPED in the BSDS.
System Action:The queue manager terminates with a dump.
System Programmer Response:The log RBA range must be available for the queue manager to be recoverable. Correct the previous errors and restore the active log data sets that contain the RBA range reported in the message.
- If the log data sets are recoverable, the active log data set inventory in the BSDS must be modified to reset the STOPPED status. Use the print log map utility (CSQJU004) to obtain a copy of the BSDS log inventory. Next, use the change log inventory utility (CSQJU003) to delete the active log data sets marked STOPPED (use the DELETE statement), then add them again (use the NEWLOG statement). The starting and ending RBA for each active log data set must be specified on the NEWLOG statement when the logs are added back to the BSDS using the change log inventory utility.
- If the log data sets are not recoverable, see the WebSphere MQ for z/OS System Administration Guide for information about dealing with problems on the log.
Problem Determination:Examine previous messages to determine the reason the active log data sets are unavailable.
FULL ARCHIVE LOG VOLUME DSNAME=dsname, STARTRBA= sss ENDRBA=ttt, STARTLRSN=ppp ENDLRSN=qqq, UNIT=unitname, COPYnVOL=vvv VOLSPAN=xxx CATLG=yyyExplanation:Offloading for the specified archive log data set was successfully completed for the given volume. If the data set spans multiple tape volumes, this message is generated for each tape volume.
This message is issued in place of CSQJ003I for queue-sharing groups.
System Action:See message CSQJ003I. STARTTIME and ENDTIME are replaced by the following:
- STARTLRSN
- The starting LRSN contained in the volume for queue-sharing groups.
- ENDLRSN
- The ending LRSN contained in the volume for queue-sharing groups.
CSQJ060Eparm-name system parameters are unusableExplanation:The format of the parameters set by parm-name in the system parameter load module is invalid, so they cannot be used.
System Action:The queue manager is terminated with abnormally with reason code X'00E80084'.
System Programmer Response:Ensure that the queue manager is started with a correct system parameter module, for example CSQZPARM. If necessary, reassemble the module that uses the indicated parameters, and relink-edit your system parameter load module.
CSQJ061Iparm-name system parameters are obsoleteExplanation:The parameters set by parm-name in the system parameter load module use some values which are now obsolete.
System Action:Processing continues. The obsolete parameters are ignored, and default values are used for new parameters.
System Programmer Response:Review your system parameter settings. If necessary, reassemble the module that uses the indicated parameters, and relink-edit your system parameter load module.
CSQJ070Ecsect-name ARCHIVE LOG DSN PREFIX NOT IN PROPER FORMAT TO RECEIVE TIME STAMP DATA. TIME STAMPING OF dsname BYPASSEDExplanation:The system parameters (set by CSQ6ARVP) specify that the date and time of creation of an archive log data set be included as part of the archive log data set name (DSN). To accomplish this, MQ requires that the length of the archive log data set name prefix is limited. If the prefix requirement is not met, this message is issued just prior to the allocation of the archive log data set specified in the message.
System Action:The archive log data set will be allocated using the archive log prefix. However, the archive log DSN will not contain the date and time as the user requested.
System Programmer Response:The system parameters for the log archive function must be changed. Specifically, the TSTAMP and ARCPFXn fields are not consistent with one another. For information about the actions required to eliminate this problem, see the WebSphere MQ for z/OS System Setup Guide, under the CSQ6ARVP macro subtopic.
CSQJ071Ecsect-name TIMER FAILURE CAUSED TIME STAMPING OF ARCHIVE dsname TO BE BYPASSEDExplanation:The system parameters (set by CSQ6ARVP) specify that the date and time of creation of an archive log data set be included as part of the archive log data set name (DSN). However an attempt to get the current date and time from the system was unsuccessful. This message is issued just prior to the allocation of the archive log data set specified in the message.
System Action:The archive log data set will be allocated using the archive log prefix. However, the archive log DSN will not contain the date and time as the user requested.
CSQJ072EARCHIVE LOG DATA SET dsname HAS BEEN ALLOCATED TO NON-TAPE DEVICE AND CATALOGED, OVERRIDING CATALOG PARAMETERExplanation:The system parameters (set by CSQ6ARVP) specify that all archive log data sets should be uncataloged (CATALOG=NO). However, MQ requires that all archive log data sets allocated to non-tape devices must be cataloged. The archive log data set specified by dsname has been allocated to a non-tape device, and has thus been cataloged. The user's system parameter CATALOG setting of NO has been overridden.
System Action:The archive log data set has been allocated to a nontape device, and has been cataloged. The system parameter CATALOG=NO setting has been overridden. The BSDS reflects that the data set has been cataloged.
System Programmer Response:The archive system parameters must be changed. Specifically, the CATALOG and UNIT parameters are not consistent with one another. For information about the actions required to eliminate this problem, see the WebSphere MQ for z/OS System Setup Guide, under the CSQ6ARVP macro subtopic.
CSQJ073ELOG ARCHIVE UNIT ALLOCATION FAILED, REASON CODE=ccc. ALLOCATION OR OFFLOAD OF ARCHIVE LOG DATA SET MAY FAILExplanation:While building the SVC99 text entries to allocate a new archive log data set dynamically, a unit allocation error was detected. The reason code, indicated by ccc in the message, further clarifies the problem as follows:
- 4-28 (X'4'-X'1C')
- Return code from z/OS IEFGB4UV macro. Common values are:
- 4 (X'04')
- Invalid unit name
- 8 (X'08')
- Unit name has incorrect units assigned
- 16 (X'10')
- No storage available
- 20 (X'14')
- Device numbers not valid
- 32 (X'20')
- MQ was able to obtain a list of devices corresponding to the device type (unit name) specified in the system parameters. However, it was determined that this list contained a mixture of tape and nontape devices.
- 36 (X'24')
- Nonfetch-protected storage could not be obtained to build a parameter list for a z/OS service.
- 40 (X'28')
- The device type (unit name) specified by the user in the system parameters is valid. However, no devices are currently associated with the given device type (unit name).
- 44 (X'2C')
- The device type (unit name) specified by the user in the system parameters is valid. However, no DASD volumes are available with a volume use attribute of storage.
System Action:This message is issued after the SVC99 text entries are built, but prior to the allocation of the new archive log data set. As a result of the error, the dynamic allocation of the archive log data set will be attempted using standard default values. The standard default values are generally acceptable; however, the allocation might be unsuccessful or the subsequent offload might produce undesirable processing results. For example:
- A reason code of 4 or 44 (X'2C') indicates an allocation error (CSQJ103E) when the SVC99 is issued for the archive data set.
- Offload processing to tape might be unsuccessful. MQ uses a volume count of 20 when allocating to tape, and uses the standard z/OS volume count default of 5 volumes when writing to non-tape devices. In the case of most of the above errors, it would be impossible for MQ to determine the device type on which the data set is to be allocated. Therefore, the standard z/OS default is assumed for the volume count. If the data set is successfully allocated to a tape device, and the volume of data is such that more than five volumes will be used for the archive data set, the offload processing will receive a z/OS completion code X'837-08' with message IEC028I when attempting to write to the sixth tape volume.
- Offload processing to a direct access device might be unsuccessful. When allocating a new archive log data set on a direct access device, MQ will use a unit count to facilitate multivolume archive data sets. With most of the above errors, it might be impossible for MQ to correctly determine the type of device on which the data set is to be allocated. Therefore, the standard default (1) is assumed for the unit count. If the data set is successfully allocated to a direct access device, and during the offload processing it becomes necessary to extend the data set to another device, the offload processing will receive a z/OS X'B37' (out of space) completion code, and the archive log data set will be deallocated.
System Programmer Response:The required action is based on the reason code indicated in the message:
- 4-28 (X'4'-X'1C')
- See the MVS Authorized Assembler Services Guide for more info about the return code from the z/OS IEFGB4UV macro. The most likely causes for the common values are:
- 4 (X'04')
- Incorrect specification in the archive system parameters. Correct the UNIT parameter. If the UNIT parameter from the archive system parameters appears to be correct, check the EDT to ensure that the esoteric or generic unit name specified in the parameters is actually in the EDT. Subsequent offload processing will archive the log data which could not be previously archived due to the allocation error (CSQJ103E).
- 8 (X'08')
- Incorrect specification in archive system parameters, incorrect operational setup.
- 16 (X'10')
- This is usually a temporary problem. If the allocation of the archive log data set is successful, no action is required to correct this situation. If this is a recurring problem, sufficient page space is not available, and the region size for the queue manager address space might have to be increased, or standard z/OS diagnostic procedures might have to be used to correct the problem.
- 20 (X'14')
- Incorrect specification in archive system parameters, incorrect operational
- 32 (X'20') or 40 (X'28')
- To correct this situation, change the archive system parameter UNIT to use a device type (unit name) that contains homogenous devices, or modify the device list associated with the device type (unit name) using a system generation to supply a list of homogenous devices.
- 44 (X'2C')
- To correct this situation, issue the z/OS command MOUNT to change the volume use attribute of a mounted private volume to storage. If this is a recurring problem, you might have to do one of the following:
- Perform a system generation to add permanently resident volumes with a volume use attribute of storage to the esoteric or generic unit
- Change the archive system parameters to use a different esoteric or generic unit name for the UNIT
CSQJ077ELOG OR BSDS READ ERROR FOR QMGR qmgr-name, REASON CODE=cccExplanation:This message identifies a queue manager whose log data cannot be accessed. The logs or BSDSs of other queue managers in a queue-sharing group may be accessed during a RECOVER CFSTRUCT operation.
System Action:The execution unit writes a record to SYS1.LOGREC and requests an SVC dump. The execution unit then terminates abnormally.
System Programmer Response:Look for earlier messages which may identify more specifically the data set being accessed and the problem.
If you are unable to solve the problem, note the reason code, collect the following items, and contact your IBM support center:
- System dump
- Console output for the issuing queue manager
- Console output for the other queue manager
- Printout of SYS1.LOGREC
CSQJ098Ecsect-name RESTART CONTROL ENDLRSN rrr IS NOT IN KNOWN LRSN RANGE. QUEUE MANAGER STARTUP IS TERMINATEDExplanation:A conditional restart control record requests truncation, but it cannot take place because the end LRSN was not in the range of LRSN values known to either the active or archive logs. rrr is the end LRSN specified in the active record. The end LRSN is either higher than the end LRSN of the most recent active log data set, or lower than the starting LRSN of the oldest archive log data set.
System Action:Queue manager startup is terminated.
System Programmer Response:Check the ENDLRSN value specified in the conditional restart control record. If it is not correct, run the change log inventory utility (CSQJU003) using CRESTART CANCEL cancel the conditional restart, and a new CRESTART specifying the correct ENDLRSN.
CSQJ099ILOG RECORDING TO COMMENCE WITH STARTRBA=sssExplanation:This message is generated during queue manager startup. The value specified by STARTRBA is the RBA of the next byte of log data to be recorded in the active log data sets.
This message is preceded by one (if single logging) or two (if dual logging) CSQJ001I messages.
System Programmer Response:None required. However, if recovery is required, information from this message might be required as input to the change log inventory utility (CSQJU003).
CSQJ100Ecsect-name ERROR OPENING BSDSn DSNAME=dsname, ERROR STATUS=eeiiExplanation:During startup, or while processing a RECOVER BSDS command, MQ could not open the specified BSDS. BSDSn matches the DDname in the queue manager started task JCL procedure (xxxxMSTR) of the data set that cannot be opened. The value of n is 1 or 2. The error status contains the VSAM open return code in ee, and the VSAM open reason code in ii.
System Action:When this error occurs at initialization time, startup must be terminated, because the log data sets cannot be determined and allocated without the BSDS. When this error occurs during RECOVER BSDS processing, the command is terminated, and the queue manager continues in single BSDS mode.
System Programmer Response:Recover the BSDS that cannot be opened. See the WebSphere MQ for z/OS System Administration Guide for information about dealing with problems on the BSDS or the log.
Problem Determination:The error status contains the VSAM open return code in ee, and the VSAM open reason code in ii. See the DFSMS/MVS Macro Instructions for Data Sets manual for a list of the VSAM OPEN return codes and reason codes, and the steps required to take corrective action.
CSQJ101Ecsect-name RESTART CONTROL ENDRBA rrr IS NOT IN KNOWN RBA RANGE. QUEUE MANAGER STARTUP IS TERMINATEDExplanation:A conditional restart control record requests truncation, but it cannot take place because the end RBA was not in the range of RBA values known to either the active or archive logs. rrr is the end RBA specified in the active record. The end RBA is either higher than the end RBA of the most recent active log data set, or lower than the starting RBA of the oldest archive log data set.
System Action:Queue manager startup is terminated.
System Programmer Response:Check the ENDRBA value specified in the conditional restart control record. If it is not correct, run the change log inventory utility (CSQJU003) using CRESTART CANCEL cancel the conditional restart, and a new CRESTART specifying the correct ENDRBA.
Otherwise, then most likely, the archive log data set that contained the requested RBA has been deleted from the BSDS data set by the change log inventory utility. Locate the output from an old print log map utility and identify the data set that contains the missing RBA. If the data set has not been reused, run the change log inventory utility to add this data set back into the inventory of log data sets. Restart the queue manager.
CSQJ102ELOG RBA CONTENT OF LOG DATA SET DSNAME=dsname, STARTRBA= sss ENDRBA=ttt, DOES NOT AGREE WITH BSDS INFORMATIONExplanation:The log RBA range shown in the BSDS for the specified data set does not agree with the content of the data set.
System Action:Startup processing is terminated.
System Programmer Response:Use the print log map and change log inventory utilities to make the BSDS consistent with the log data sets.
CSQJ103Ecsect-name LOG ALLOCATION ERROR DSNAME=dsname, ERROR STATUS=eeeeiiii, SMS REASON CODE=ssssssssExplanation:An error occurred while attempting to allocate the active or archive log data set indicated by DSNAME. STATUS indicates the error reason code returned by z/OS dynamic allocation (SVC99).
This message might be preceded by message CSQJ073E.
System Action:Subsequent actions depend on the type of data set involved.
For active log data sets, if the error is encountered during queue manager initialization, startup is terminated. If two copies of the active log data sets are defined, this message appears only once.
For archive log data sets, if two copies of the archive log data sets are defined, processing continues on the remaining archive log data set.
System Programmer Response:The error status portion of this message contains a 2-byte error code (eeee, S99ERROR) followed by the 2-byte information code (iiii, S99INFO) from the SVC99 request block. If the S99ERROR code indicates an SMS allocation error ('97xx'), then ssssssss contains additional SMS reason code information obtained from S99ERSN. See the MVS Authorized Assembler Services Guide manual for a description of these codes.
For active log data sets, if the problem occurred during queue manager initialization, we can resolve the problem by doing one of the following:
- Resolve the error associated with the active log data set as indicated by STATUS
- Provide another copy of the active log data set, using Access Method Services
- Update the BSDS with the change log inventory utility (CSQJU003)
- Restart the queue manager
- If the problem occurred during allocation with the intent to write the data set, no immediate action is required. However, if you do not resolve the SVC99 error (indicated by the STATUS value in the message), the available space in the active log could eventually be exhausted (CSQJ111A) because all future offloads might be unsuccessful because of the same error.
- If the problem occurred during allocation with the intent to read the data set, determine the problem, and use the change log inventory utility (CSQJU003) DELETE function to delete the archive log data set from the BSDS archive log inventory. Then use the NEWLOG function to add the data set back into the archive log inventory, pointing to the correct volume and device.
See the WebSphere MQ for z/OS System Administration Guide for information about dealing with problems on the log.
This message might also be issued as the result of a user error. If STATUS displays a value of '17080000', you might have one or more active log data sets defined in the BSDS, but not allocated on DASD. To correct the situation, print the contents of the current active log data set inventory using the print log map utility (CSQJU004), then either:
- Use Access Method Services to allocate the active log data set for each active log data set listed in the BSDS, but not actually allocated on DASD. We can find the Access Method Services command syntax for active log data sets in the CSQ4BSDS sample JCL.
- Use the change log inventory utility (CSQJU003) DELETE statement to delete the errant active log data set name, and the NEWLOG statement to add the correct name to the active log data set inventory. The name specified on the NEWLOG statement must be the same as the name of the actual active log data set allocated on DASD.
CSQJ104Ecsect-name RECEIVED ERROR STATUS nnn FROM macro-name FOR DSNAME dsnameExplanation:An error ocurred while issuing macro macro-name. Error status is the return code from the specified macro:
- For an OPEN of a VSAM data set, the return code in the error field of the Access Method Services control block is included in this message as the error status value. See the DFSMS/MVS Macro Instructions for Data Sets manual for a description of these values.
- If the OPEN was for a non-VSAM data set, the error status is zero.
- For MMSRV errors, error status contains the error information returned by media manager services. If an MMSRV CATUPDT error occurs attempting to truncate an active log data set, the log data set will be unavailable and the status of the log data set will be flagged as STOPPED in the BSDS.
- For VSAM OPEN and MMSRV errors, this message is preceded by an IEC161I message that defines the error that occurred.
- For a PROTECT of an archive log data set, the return code is from DADSM PROTECT. See the MVS/ESA System - Data Administration manual for details of the return code.
See the WebSphere MQ for z/OS System Administration Guide for information about dealing with problems on the log.
System Action:Subsequent actions depend on when the error occurred.
During queue manager initialization, startup is terminated.
When using the data set either for offload or for input operations, processing continues. If a second copy of the data is available, MQ attempts to allocate and open the second data set.
When using the data set as an active log data set, MQ attempts to retry the request. If the retry is unsuccessful, the queue manager is terminated.
During checkpoint processing, where MQ attempts to locate the oldest active or archive log data sets that are required for restart recovery of page sets and restart and media recovery of CF structures, processing continues. The message is a warning that either restart recovery would fail or media recovery of CF structures would fail. It is most likely to occur when all CF application structures are not being regularly backed up, thereby requiring excessively old log data sets for recovery.
System Programmer Response:If the error occurred during initialization, either correct the problem so that the data set is available or provide another copy of the data set and change the BSDSs to point to the new data set.
If the error occurred after startup, the return code should be reviewed and the appropriate action taken to correct the problem, so that the data set can be used at a later time, or the data set entry can be removed from the BSDS using the change log inventory utility.
If the error was received from PROTECT, there might be a problem with the PASSWORD data set. See the appropriate DADSM publication to determine the cause of the problem. When the problem has been corrected, ensure the archive log data sets receiving the error are added to the PASSWORD data set. If these archive log data sets are not added to the PASSWORD data set, archive read will not be able to OPEN these data sets. If you do not have information about the named macro, note the macro name and the return code and contact your IBM support center for help.
If the error occured during checkpoint processing, issue the DISPLAY USAGE TYPE(DATASET) command to show which log data sets are currently required for page set and media recovery, and ensure that they are available. If applicable, use the BACKUP CFSTRUCT command for your CF structures, and institute a procedure to back up your CF structures frequently.
CSQJ105Ecsect-name LOG WRITE ERROR DSNAME=dsname, LOGRBA=rrr, ERROR STATUS=ccccffssExplanation:An error occurred while writing a log data set. If csect-name is CSQJW107, the error occurred writing the log buffers to an active log data set. If csect-name is CSQJW207, the error occurred while preformatting the next control area before writing log data into it.
Error status contains the error information returned by media manager in the form ccccffss, where cccc is a 2-byte return code that describes the error, ff is a 1-byte code that defines the functional routine that detected the error, and ss is the 1-byte status code that defines a general category of error.
System Action:If the dual active logging option is selected, the MQ switches to the next data set for this copy. If the next data set is not ready, MQ temporarily enters single logging mode and allocates a replacement data set for the one that encountered the error. Dual logging is resumed as soon as possible.
If single active logging option is selected and the next data set is not ready, MQ waits for that data set to be available. In this case, log writing is inhibited until the replacement is ready for output.
System Programmer Response:See the MVS/DFP Diagnosis Reference manual for information about return codes from the media manager. If you are unable to resolve the problem, note the return code, and contact your IBM support center.
CSQJ106ELOG READ ERROR DSNAME=dsname, LOGRBA=rrr, ERROR STATUS=ccccffssExplanation:An error occurred while reading an active log data set. The error status contains the error information returned by the media manager in the form ccccffss, where cccc is a 2-byte return code that describes the error, ff is a 1-byte code that defines the functional routine that detected the error, and ss is the 1-byte status code that defines a general category of error. (See the MVS/DFP Diagnosis Reference manual for information about return codes from the media manager.)
System Action:If another log data set contains the data, MQ attempts to read the data from the alternate source. If an alternate source is not available, a read error return code is sent to the program requesting the log data. Depending on the circumstances under which the failure occurred, the queue manager might continue with the alternate log data set if dual logging is used, or end abnormally.
System Programmer Response:If you are using dual logging, the requested RBA was probably retrieved from the corresponding dual active log data set, and no immediate response is necessary. However, if this error occurs frequently, or if you are using single logging, immediate attention might be required. If so, note the contents of the error status field, and contact your IBM support center for help.
It might be necessary to replace the data set in error with a new data set containing the log data, and to update the BSDSs to reflect the new data set using the change log inventory (CSQJU003) NEWLOG operation.
See the WebSphere MQ for z/OS System Administration Guide for information about dealing with problems on the log.
This message might also be issued as the result of a user error. If the data set name specified by DSNAME is missing, and STATUS displays a value of '00180408' or '00100408', you are using dual logging, but only one set of active log data sets is defined in the BSDS. To resolve this condition, do either of the following:
- Define a second set of active log data sets using Access Method Services (if they are not defined already), and update the BSDS log inventory using the change log inventory (CSQJU003) NEWLOG operation. See the WebSphere MQ for z/OS System Administration Guide for information about using the change log inventory utility.
- Reset the log system parameters to indicate single logging. We can do this by setting TWOACTV to 'NO' in the CSQ6LOGP system parameters.
CSQJ107EREAD ERROR ON BSDS DSNAME=dsname ERROR STATUS=eeeExplanation:An error occurred while reading the specified BSDS. Error Status contains the VSAM return and feedback codes. It is a 2-byte field with the first byte containing the hexadecimal return code and the second containing the hexadecimal feedback code. See the DFSMS/MVS Macro Instructions for Data Sets manual for a description of VSAM return and reason codes.
See the WebSphere MQ for z/OS System Administration Guide for information about dealing with problems on the BSDS or the log.
System Action:If dual BSDSs are available, MQ attempts to read from the other BSDSs. If the read from the second BSDS fails or if there is only one BSDS, an error code is returned to the log request that caused access to the BSDS.
If the read error is detected during startup, the queue manager terminates.
System Programmer Response:It might be necessary to replace or repair the BSDS, depending on what conditions resulted from the read error. To replace a BSDS, first delete the BSDS in error, then define the new BSDS with the same name and attributes. If a new name is used for the new BSDS, change the queue manager started task JCL procedure (xxxxMSTR) to specify the new BSDS name.
CSQJ108EWRITE ERROR ON BSDS DSNAME=dsname ERROR STATUS=eeeExplanation:An error occurred while writing to the specified BSDS. Error Status contains the VSAM return and feedback codes. It is a 2-byte field with the first containing the hexadecimal return code and the second containing the hexadecimal feedback code. See the DFSMS/MVS Macro Instructions for Data Sets manual for a description of VSAM return and reason codes.
System Action:If dual BSDSs are available, MQ enters single BSDS mode using the remaining good BSDS. Otherwise, an error code is returned to the log request that caused access to the BSDS.
System Programmer Response:If dual BSDS mode is being used, run an offline Access Method Services job to rename the error BSDS and define a new BSDS with the same name. Then enter the RECOVER BSDS command to reestablish dual BSDS mode.
If dual BSDS mode is not being used, the queue manager must be shut down, and the BSDS must be recovered from a backup copy. To recover the BSDS, use the change log inventory utility.
CSQJ109EOUT OF SPACE IN BSDS DSNAME=dsnameExplanation:There is no more space in the specified BSDS. The operation that encountered the out-of-space condition did not complete properly.
System Action:If dual BSDSs are available, MQ enters single BSDS mode using the remaining good BSDS. Otherwise, an error code is returned to the log request that caused access to the BSDS.
System Programmer Response:If dual BSDS mode is being used, run an offline Access Method Services job to rename the full BSDS and define a new, larger BSDS with the same name. Enter the RECOVER BSDS command to reestablish dual BSDS mode.
If dual BSDS mode is not being used, the queue manager must be shut down and the BSDS recovered offline. In this case, run the same Access Method Services job mentioned above to rename the full data set and define a larger data set. Next, run an Access Method Services REPRO job to copy the full BSDS into the new BSDS.
CSQJ110ELAST COPYn ACTIVE LOG DATA SET IS nnn PERCENT FULLExplanation:This message is issued when the last available active log data set is 5% full, and is reissued after each additional 5% of the data set space is filled.
System Action:Each time the message is issued, the offload processing will be re-attempted. If the situation is not corrected, the active log data set will fill to capacity, message CSQJ111A will be issued, and MQ processing will stop.
System Programmer Response:To clear this condition, take steps to complete other waiting offload tasks. Once an active log data set is made available (reusable) by completing the offload process for it, the MQ logging activity can continue.
Perform a display request to determine the outstanding requests related to the log offload process. Take the necessary action to satisfy any requests, and permit offload to continue.
Consider whether there are sufficient active log data sets. If necessary, additional log data sets can be added dynamically using the DEFINE LOG command.
If offload does not complete normally or cannot be initiated, either correct the problem that is causing the offload process error, increase the size of the allocated data sets, or add more active log data sets. Note that the latter action requires the queue manager to be inactive and the change log inventory utility to be run.
Possible causes for the shortage of active log data space are:
- Excessive logging. For example, there is a lot of persistent message activity.
- Delayed or slow offloading. For example, failure to mount archive volumes, incorrect replies to offload messages, or slow device speeds.
- Excessive use of the ARCHIVE LOG command. Each invocation of this command causes MQ to switch to a new active log data set and to initiate an offload of the active log. Although the command will not be processed when only one active log data set remains in a copy of the active log (see CSQJ319I), excessive use of the command could have consumed all space in the active log except the current active log data sets.
- Offloads were unsuccessful.
- Insufficient active log space.
CSQJ111AOUT OF SPACE IN ACTIVE LOG DATA SETSExplanation:Due to delays in offload processing, all available space in all active log data sets has been exhausted. Recovery logging cannot continue.
System Action:MQ waits for an available data set. Any tasks performing MQ API calls that require logging will wait.
System Programmer Response:Perform a display request to ensure that there are no outstanding requests that are related to the log offload process. Take the necessary action to satisfy any requests, and permit offload to continue.
Consider whether there are sufficient active log data sets. If necessary, additional log data sets can be added dynamically using the