Backup procedures

You need not back up the JCL repository (JS) data set on a daily basis for recovery purposes. HCL Workload Automation for Z automatically backs up the JS file based on the value specified for the MAXJSFILE keyword of the JTOPTS statement for the subsystem. However, you can disable this automatic backup and schedule JS file backups as job operations using the BACKUP command; for example, if you want to schedule backups during times when the workload on the system is low. The JS data set consists of two data sets, one active, the other inactive. Recovery is simply a matter of copying the inactive data set to the active data set.

You can back up VSAM data sets using the REPRO function of the IDCAMS utility program. But you cannot use REPRO if the backup is to a sequential file and the record length is greater than 32760. Instead, use DFSMSdss™, or an equivalent product, to perform the backup. If you use DFSMS, consider DFSMShsm™ ABARS for data backup and restore. ABARS simplifies the backup and recovery process.

Normally, VSAM data sets are unloaded to a sequential data set. A recommended practice is to use a generation data group as the backup data set. If backup data sets are allocated on DASD, they must be on a different volume from the VSAM data set being backed up. On a 3380 direct access storage device, the backup data set should be on a different head disk assembly from the data set being backed up.

The long-term plan data set principally provides input for the daily plan batch jobs that create a new current plan. Some of the daily plan batch jobs update the long-term plan data set to set the status of occurrences in the long-term plan. The long-term plan and the current plans must synchronize otherwise daily-plan processing will fail.

HCL Workload Automation for Z creates a backup of the long-term plan when long-term plan and daily-planning batch jobs are run. The backup is written to a VSAM file with the ddname EQQLTBKP.

Before and after the new long-term plan is generated, a backup is made for long-term batch jobs. This ensures that a usable long-term plan is available in the backup file with the ddname EQQLTBKP after a long-term plan batch run.

For data processing batch jobs, a long-term backup is made after the new current plan has been successfully created. This ensures that exists a long-term plan backup that is synchronized with the new current plan produced in the file with ddname EQQNCPDS.