Remote Queue Utility (FDRABRUT)

Your installation may decide to allow end-users to request ABR processing of the data sets they own. Although it is possible for a user to submit their own ABR job, it is wasteful to create an incremental backup or Archive Backup file for just a few data sets. Restores can also be made more efficient if the requests are “batched” together, minimizing tape mounts.

As noted in FDRABR-Volume-Backups and elsewhere in FDRABR-Archiving-and-Superscratch, ABR supports optional remote queues. These queues allow users to request certain ABR functions to be performed on data sets they specify. The remote queues are not processed immediately. The requests are handled when an appropriate ABR batch job is run by data center operations or a scheduling program.

The remote queues are managed by the utility program FDRABRUT and can be used for four types of requests:

  1. Include a data set in ABR Volume Backups
  2. Archive a data set
  3. Restore a data set from Volume Backups
  4. Restore a data set from Archive Backups.

You can support all four types of requests, or any subset of them.

For all four request types, a remote queue data set can be used. These are simple sequential data sets, and control statements specifying the user's request are added to the end of them. There is a separate remote queue data set for each type of request. Users can use one of several techniques to add requests to a queue (all involving FDRABRUT) and an appropriate ABR job (backup, archive, restore) reads and processes the requests. Since it is possible that users are adding requests at the same time that ABR is reading them, an internal enqueue is used to serialize access to each remote queue data set. ABR empties the data set when the requests are processed.

For the first two request types (backup and archive), you can optionally set a flag in the Format 1 DSCB of the data set to be processed. For Volume Backup, it simply sets the update flag that will cause it to be included in the next incremental backup. For Archive, it sets a flag in the reserved byte used by ABR, causing it to be archived. These are effective only if their DASD volume is processed in the next appropriate ABR run. This option is used if DISKUPDATE=YES is set in the FDR Global Options (see “DISKUPDATE” in Section 90.8) but it works only if FDRABRUT can be executed as an authorized program from an authorized library.

TSO users

TSO users have several ways of executing FDRABRUT to add requests to a remote queue:

  • They can execute it directly, for example, CALL lib(FDRABRUT) and interactively enter the control statements documented in the following sections.
  • Many ABR ISPF panels have a “remote queue” option, to add the request to a queue. FDRABRUT statements are generated by the panel and the utility is dynamically invoked.
  • There is an ABF ISPF panel devoted to remote queue operations of all types. FDRABRUT statements are generated by the panel and the utility is dynamically invoked.
  • When a TSO user references an archive data set, one of the options they may be presented is to add the request to a queue (RQ). FDRABRUT statements are generated by the ABR Catalog Locate exit and the utility is dynamically invoked.

If the request is for backup or archive, and the DISKUPDATE=YES option is set in the FDR Global Options, FDRABRUT fails unless it is executed as an authorized program under TSO. See Authorize-the-FDR-Programs for information on authorizing programs under TSO.

Batch jobs

Batch jobs can execute FDRABRUT directly, using the JCL and control statements documented in the following sections. This might be used to request that a newly created data set be archived almost immediately, by putting an FDRABRUT step in the creating job.

Security

If the ALLCALL option is enabled in the FDR Global Options (see ALLCALL in Security-Options; FDRABRUT does not yet support the FDR.NOALLCALL security profile), FDR security checking is enabled and users are only able to add requests to the various remote queues if they have a proper level of authority to each data set being added. FDRABRUT checks for the proper authority before adding the request.

If DISKUPDATE=YES is in effect for backup requests, FDRABRUT will follow the rules defined in Security in Processing-Options-and-Requirements (for backup) and Security in FDRABR-Processing-Options-and-Requirements (for archive). However, if DISKUPDATE=NO, FDRABRUT will not do a volume authority check; it will only check for authority to the individual data sets

For backup requests when DISKUPDATE=NO is in effect, and for all restore requests, the use of DSG= to select groups of data sets is disabled when security is enabled, since FDRABRUT is not able to determine the names of the individual data sets for security checks. Only DSN= with a fully-qualified data set name is supported.

The remote queue data sets

Create-the-Remote-Queue-Files has details on creating and identifying the 4 remote queue data sets. You may choose to implement any or all of them.

FDRABRUT dynamically allocates the proper remote queue data set for each control statement, so the data set names of the remote queues must be specified in the FDR Global Options (panel A.I.4.5). There is a default name for each data set, but you can change it to anything you prefer. Panel A.I.4.5 – Set FDR Global Data Set Name Options in ABR-Options shows the default names.

Servicing the remote queues

An appropriate ABR job must be run periodically to process any requests on each of the queues.

For the backup and archive queues, they can be included in your regular daily Volume Backup or Archive Backup ABR jobs. The requested data sets are simply included with those that are selected for other reasons.

For the two restore queues, executing a periodic job of the proper type (RESTORE TYPE=ABR or TYPE=ARC) that restores only the selected data sets is needed. In order to meet user's expectations, this job should be run at regular intervals (for example, once an hour, once every 4 hours).

If DISKUPDATE=YES is in effect for the backup and archive queues, data sets on the queue are processed the next time that an ABR backup or archive job processes the volumes containing those data sets. Data sets on the backup queue are automatically included in the next incremental, even if they have not been updated, while data sets on the Archive queue may not be archived unless you specify the REMOTE operand on the DUMP TYPE=ARC statement.

For the two restore queues, and the backup and archive queues if DISKUPDATE=NO is in effect, use of the remote queue for a particular ABR function is controlled by the presence of a particular ddname in the JCL of the ABR step. Here are the ddnames and the functions that invoke them:

ABRBKDQ

DUMP TYPE=ABR/AUTO/DSF

ABRARDQ

DUMP TYPE=ARC

ABRREST

RESTORE TYPE=ABR

ABRARCH

RESTORE TYPE=ARC

For more details, see:

 

Tip: For faster searching, add an asterisk to the end of your partial query. Example: cert*