How do you set up backup and recovery for CA Workload Automation EE files?
Backup and recovery can be done for the following key files:
The backup datasets can be defined in init parm. The scheduled times that these backups run depends on the file as listed above. The files and the commands that set the backup schedule are listed below. To start the backups, each file has a command that starts with BKUP. For example, BKUPEVS to backup the event data set. In addition, the history file also can be scheduled for a backup as defined in the init parm.
In addition to static backups, the product also can record all changes to the files. So if backup is done on Friday and then recovered on Sunday, all changes to the files can be applied. This is called JOURNALING. Each change to a file is recorded in an SMF record and these records are read in when a recovery is done.
This is how journaling is set up:
- In the INIT parms add a SMFREC record for the SMF record number to use. For example:
This will use SMF record 200 for journaling of file changes. This SMF record number must be unique.
- Then in each of the files being backed up, the JOURNAL option should be specified in their init parm.
To restore a dataset, issue the restore command. The following is an example:
//STEP01 EXEC PGM=ESP,PARM='SUBSYS(ESPM)'
//STEPLIB DD DSN=CYB2.ESP510.SSCPLINK,DISP=SHR
//SYSPRINT DD SYSOUT=*
//JOURNAL DD DSN=SYS1.SMF.BACKUP(0),DISP=SHR
// DD DSN=SYS1.SMF.BACKUP(-1),DISP=SHR
//SYSIN DD *
RESTORE HISTFILE('CYB2.ESP510.HISTFILE') +
In the above example, note that the DD JOURNAL point to the datasets that contains SMF records. The SMF records in the file must be for the time period between the backup and the restore(all the updates).
Perform the above for each file that was backed up. For the remaining files, they need to be restored from pack backups(DFDSS) or reformatted. This is all dependent on the file.