Changes for active logs

Use this topic to understand how you can change the active logs using the BSDS.

You can add to, delete from, and record entries in the BSDS for active logs using the change log utility. Examples only are shown here; replace the data set names shown with the ones you want to use. For more details of the utility, see The change log inventory utility.

See these sections for more information:
Adding record entries to the BSDS
If an active log has been flagged as "stopped", it is not reused for logging; however, it continues to be used for reading. Use the access method services to define new active log data sets, then use the change log inventory utility to register the new data sets in the BSDS. For example, use: 

NEWLOG DSNAME=MQM111.LOGCOPY1.DS10,COPY1
NEWLOG DSNAME=MQM111.LOGCOPY2.DS10,COPY2

If you are copying the contents of an old active log data set to the new one, you can also give the RBA range and the starting and ending time stamps on the NEWLOG function.

Deleting information about the active log data set from the BSDS
To delete information about an active log data set from the BSDS, you could use: 

DELETE DSNAME=MQM111.LOGCOPY1.DS99
DELETE DSNAME=MQM111.LOGCOPY2.DS99
Recording information about the log data set in the BSDS
To record information about an existing active log data set in the BSDS, use: 

NEWLOG DSNAME=MQM111.LOGCOPY1.DS10,COPY2,STARTIME=19930212205198,
ENDTIME=19930412205200,STARTRBA=6400,ENDRBA=94FF
You might need to insert a record containing this type of information in the BSDS because:
  • The entry for the data set has been deleted, but is needed again.
  • You are copying the contents of one active log data set to another data set.
  • You are recovering the BSDS from a backup copy.
Increasing the size of the active log
There are two methods of achieving this process.
  1. When the queue manager is active:
    1. Define new larger log data sets using JCL.
    2. Add the new log data sets to the active queue manager using the MQSC DEFINE LOG command.
    3. Use the MQSC ARCHIVE LOG command to move the current active log, to be a new larger log.
    4. Wait for the archive of the smaller active log dataset to complete.
    5. Shutdown the queue manager, using the CSQJU003 utility to remove the old small active logs.
    6. Restart the queue manager.
  2. When the queue manager is inactive:
    1. Stop the queue manager. This step is required because IBM® MQ allocates all active log data sets for its exclusive use when it is active.
    2. Use Access Method Services ALTER with the NEWNAME option to rename your active log data sets.
    3. Use Access Method Services DEFINE to define larger active log data sets.

      By reusing the old data set names, you do not have to run the change log inventory utility to establish new names in the BSDSs. The old data set names and the correct RBA ranges are already in the BSDSs.

    4. Use Access Method Services REPRO to copy the old (renamed) data sets into their appropriate new data sets.
      Note: This step can take a long time, so your enterprise could be out of action for this period.
    5. Start the queue manager.

If all your log data sets are the same size, your system will be operationally more consistent and efficient. If the log data sets are not the same size, it is more difficult to track your system's logs, and so space can be wasted.

The use of CSQJUFMT
Do not run a CSQJUFMT format when increasing the size of an active log.
If you run CSQJUFMT (in order to provide a performance advantage the first time the queue manager writes to the new active log) you receive messages: 

IEC070I 203-204,XS95GTLX,REPRO02,OUTPUT,B857,SPMG02, 358
IEC070I MG.W.MG4E.LOGCOPY1.DS02,MG.W.MG4E.LOGCOPY1.DS02.DATA,
IDC3302I ACTION ERROR ON MG.W.MG4E.LOGCOPY1.DS02
IDC3351I ** VSAM I/O RETURN CODE IS 28 - RPLFDBWD = X'2908001C'
IDC31467I MAXIMUM ERROR LIMIT REACHED.

IDC0005I NUMBER OF RECORDS PROCESSED WAS 0

In addition, if you use the Access Method Services REPRO, ensure that you define a new empty log.

If you use REPRO to copy the old (renamed) data set into its respective new data set, the default is NOREPLACE.

This means that REPRO does not replace a record that is already on the designated data set. When formatting is done on the data set, the RBA value is reset. The net result is a data set that is not empty after formatting.