BRMS *SAVSYS BYPASSED for a Control Group
Have you ever seen or gotten a BRM1744 during a BRMS full system save or SAVSYSBRM? Here is what you need to do to resolve this, and I am only going to address V5R3M0 and above. When a SAVSYS is performed through BRMS using a backup control group or the SAVSYSBRM command, BRMS will automatically attempt to end the system to a restricted condition by issuing the following command:
ENDSBS SBS(*ALL) OPTION(*CNTRLD)
In V5R3M0 and above, once the time limit is up on the *CNTRLD ENDSBS SBS(*ALL), BRMS will use the new system value QENDJOBLMT to wait longer, as the system will do the aborts and clean up all the subsystem jobs, so the system will be able to get a better restricted state than previous releases.
Update this system value to allow BRMS to wait longer for the system to reach a restricted state. The system comes shipped with 120 seconds. To know what value to use, see when the ENDSBS SBS(*ALL) (CPF0994 is issued) starts and when the System ended to restricted condition (CPF0968 is issued). You can see this by running the DSPLOG command for the timeframe of your last full system save or SAVSYS. For this example, DSPLOG PERIOD((*AVAIL 081719) (*AVAIL 081819)) MSGID(CPF0994 CPF0968), the log contents are shown below:
F1 on each of the log entries, you will see the time sent for each, which will give you the time it took from when the ENDSBS command was initiated and when the system ended to restricted condition.
In this example, ENDSBS started at 14:36:24, and the system reached restricted state at 14:36:55, so well within the 120-second default value. But, if you are receiving a BRM1744, you are probably exceeding the 120-second default value.
Note: when the BRM1744 occurs, the controlling subsystem will not be started by BRMS unless the job is being run in the controlling subsystem, SBMJOB(*CTLSBS) parameter of the STRBKUBRM command.