Running automatic functions concurrently

Although the example system that has been used to explain the automatic functions of DFSMShsm used different times for automatic backup, dump, and migration, any or all of the functions can run concurrently, either in the same host or in different hosts.

When running the automatic functions concurrently, you must remember that the first thing that automatic backup does on the primary host is to optionally back up the control data sets. While the control data sets are being backed up, automatic dump, automatic volume space management and automatic secondary space management are delayed. Also, if automatic dump, automatic volume space management or automatic secondary space management are running when automatic backup begins to back up the control data sets, backup of the control data sets is delayed until migration and dump tasks complete the volumes they are processing and automatic secondary space management completes one of its level functions. If automatic backup starts first, none of the other processes can start until the control data sets are backed up.

Note: Functions such as recycle, tape copy, and audit perform similarly to automatic functions in regards to contention with CDS backup.

Example: In the example system, assume that DFSMShsm host 1 performs level functions and backup; DFSMShsm host 2 performs migration; and DFSMShsm host 3 performs dump. The commands provided are only those that would be particularly different from the examples that have been shown before.

First, the time when each of the DFSMShsm hosts begins processing must be defined. Issue the following commands:

For DFSMShsm host 1:
SETSYS AUTOBACKUPSTART(0001,0100,0200)
 
SETSYS SECONDARYSPMGMTSTART(0800,1600)
For DFSMShsm host 2:
SETSYS PRIMARYSPMGMTSTART(0030,0500)
For DFSMShsm host 3:
SETSYS AUTODUMPSTART(0030,0300,0500)

Automatic backup is started early in z/OS® image 1 to allow it to back up the control data sets. After the control data sets are backed up, the other automatic functions are started.

For concurrent processing as it has been previously defined, the z/OS images that are to process each function must be defined to the storage groups. The MIGRATE SYSTEM/SYS GROUP NAME, BACKUP SYSTEM/SYS GROUP NAME, and DUMP SYSTEM/SYS GROUP NAME attributes are used in the storage groups to provide the definitions for SMS-managed volumes.

For example, the attributes in all storage groups would be set as follows:

  BACKUP SYSTEM/SYS GROUP NAME  ===> PRC1

  MIGRATE SYSTEM/SYS GROUP NAME ===> PRC2

  DUMP SYSTEM/SYS GROUP NAME    ===> PRC3

Note: System group names can also be used for BACKUP SYSTEM/SYS GROUP NAME, MIGRATE SYSTEM/SYS GROUP NAME, and DUMP SYSTEM/SYS GROUP NAME.

Use the ADDVOL command to provide z/OS image definitions for non-SMS-managed volumes. For example, the following ADDVOL commands can be used to control the common volumes (GP0001, GP0002, GP0003).

For DFSMShsm host 1:
ADDVOL GP0001 PRIMARY(NOAUTOMIGRATION, AUTOBACKUP, -
   NOAUTODUMP)
 
ADDVOL GP0002 PRIMARY(NOAUTOMIGRATION, AUTOBACKUP, -
   NOAUTODUMP)
 
ADDVOL GP0003 PRIMARY(NOAUTOMIGRATION, AUTOBACKUP, -
   NOAUTODUMP)
For DFSMShsm host 2:
ADDVOL GP0001 PRIMARY(AUTOMIGRATION, NOAUTOBACKUP, -
   NOAUTODUMP)
 
ADDVOL GP0002 PRIMARY(AUTOMIGRATION, NOAUTOBACKUP, -
   NOAUTODUMP)
 
ADDVOL GP0003 PRIMARY(AUTOMIGRATION, NOAUTOBACKUP, -
   NOAUTODUMP)
For DFSMShsm host 3:
ADDVOL GP0001 PRIMARY(NOAUTOMIGRATION, NOAUTOBACKUP, -
   AUTODUMP)
 
ADDVOL GP0002 PRIMARY(NOAUTOMIGRATION, NOAUTOBACKUP, -
   AUTODUMP)
 
ADDVOL GP0003 PRIMARY(NOAUTOMIGRATION, NOAUTOBACKUP, -
   AUTODUMP)