Phase 2 — Performing disaster recovery

This section describes how to perform disaster recovery in a real or simulated scenario. You perform the procedures in this section at the disaster recovery site (or at the home site in your disaster recovery testing environment) after the OS/390 system resources, including user catalogs, have been restored and the OS/390 system has been started successfully.

The preparations discussed in Phase 1 are essential to the success of Phase 2; therefore, make sure that all steps in the Phase 1 procedures have been performed before you attempt Phase 2 procedures.

Performing disaster recovery consists of the following procedures:

Restoring the RUV repository and starting RUV

To restore the RUV repository, perform the following steps:

1. Define the RUV repository, and start the RUV subsystem.

2. Restore the RUV repository with the following statement:

   RESTORE REPOSITORY BACKUP_IN(DDIN);
3. Register any RUV backups or archives that have been created since the time of the repository backup.
   • If you need to change volume serial numbers or SMS classes for VSAM files, define empty VSAM files as necessary and restore the files. RUV will use current definitions as found at the remote site. This makes it fairly simple to change the volume serial numbers or SMS-managed classes.
   • If you do not need to change volume serial numbers or SMS classes for VSAM files, delete the VSAM files to be recovered from the disaster site catalogs.
   • If recovery is to be to a particular point in time, modify the RUV command to reflect the correct recovery point. When no point in time is specified, RUV will recover the files to the most current time possible.
4. Run the restore or recovery jobs as required. RUV will automatically restore files if they are not found in the z/OS catalog.

Understanding the potential for lost data after recovery

The following information may be lost when a repository is restored:

• Any changes made since the repository backup
• The value of the <+1> variable
• Registrations of subsequent repository backups

To restore this information to current using RUV, you must journal the repository changes to z/OS system management facility (SMF) logs and use the RECOVER REPOSITORY FORWARD command. See Recovering-a-RUV-repository-using-SMF-data.

Alternatively, you may manually register archives and backups that occurred after the repository backup. You must use the SET <+1>(nn) command to change the <+1> variable to the desired value.