Recovering lost data
This topic describes how to recover lost data.
About this task
The capability to resume business operations if a product or site failure is provided by the grid
configuration of the TS7700.
In a grid configuration, two or three TS7700 clusters are interconnected and
can replicate data that is created on either TS7700 cluster to the other. Data can
be restored if the recovery data is contained entirely on or through a mixture of:
- Copy export physical volumes
- Virtual volumes within a TS7700
- Physical tape
- Replicated disk
When you restore a library environment, it is important to consider and plan for:
- Restoration of the tape management system database
- Restoration of system catalogues, including the tape configuration database (TCDB) (the VOLCAT)
- Restoration of the key management environment for encrypted cartridges
- Location of input/output definition file (IODF) for devices and the LIBRARY-ID and LIBPORT-ID of the recovery TS7700
- Location of the storage management subsystem (SMS) control data set (SCDS) containing library definitions of the recovery TS7700, and construct and automatic class selection (ACS) routine definitions
- Activation of IODF and SCDS and bringing the library and devices online
- Insertion of virtual volumes that are needed for new allocations
- Restoration of the system environment and required application data
Note: Virtual volumes and data set names that are needed during the recovery process must be known.
If copy export was used, a physical volume from the last copy export set must be known since it
contains the latest copy of the internal TS7700 database. This copy describes the most recent and
all previous copy export sets. Contact your IBM Service Representative for additional
information.
You can Use DFSMSdss to dump recovery data
at the production site to virtual volumes in a TS7700.
When
no z/OS environment exists:
- Use DSS Stand-Alone Services to restore system-related data.
- Use Standard DFSMSdss RESTORE to restore other dumped data when the rescue system has been restored.
- Use RESTORE to recover dumped data. Note: TCBD and SCDS are required to mount initial recovery volumes in the TS7700 since production TCBD and SCDS can reside on virtual volumes in the recovery TS7700.
- Use ISMF to define recovery library definitions and any required constructs for a base SMS configuration if SCDS is NULL in the starter system.
- Activate SCDS.
- Use IDCAMS to create volume records for recovery volumes in the TCDB if the TCBD VOLCAT in the starter system is empty.
- Verify OAM is configured.
- Use disk replication to replicate system catalogues, TCBD, SCDS, tape management system database, and other databases and data sets.
- Determine recovery point objectives and whether asynchronous or synchronous disk replication were used.
- Use TS7700 copy policies to replicate other production-related data to the DR site.
Disaster
recovery environments can assume many configurations:
- Multi-cluster TS7700 Grid
- The grid contains two or more TS7700 systems. Virtual volumes contain the replicated recovery data.
- Empty TS7770T or TS7760T
- Physical volumes that contain the replicated recovery data and that have been copy exported are restored to the empty cluster.
As part of a total systems design, you must develop business continuity procedures that instruct I/T personnel what actions to take when a failure occurs. The best time to test the procedures is during an initial installation of the system or at some interval.
For more information about recovering lost data, see IBM® TS7700 Series Grid Failover Scenarios at http://www-03.ibm.com/support/techdocs/atsmastr.nsf/Web/AdvSearch.