Backup and Recovery product considerations

APF-authorized environment

The utility must be able to execute briefly as an authorized program at various times during processing.

Therefore, the utility attempts to establish an APF-authorized environment as follows:

The utility determines whether the product modules are being loaded from an APF-authorized STEPLIB, JOBLIB, or LNKLST concatenation of libraries:
- If the libraries are APF authorized, the utility obtains authorization through conventional methods.
- If the libraries are not APF authorized, the utility proceeds to step 2.
The utility determines whether the DBUSSID keyword is specified on the GLBL control statement or in the global options module:
- If the keyword specifies the subsystem ID of a Database Utilities Subsystem (DBUSS) that is active in the environment, the utility obtains authorization through that DBUSS.
- If the keyword is not specified on the GLBL control statement or in the global options module, the utility proceeds to Step 3.
- If the keyword specifies an inactive DBUSS, the utility proceeds to step 4.
For more information about the DBUSS, see DBUSSID keyword.
The utility determines whether the default DBUSS (subsystem ID DBUZ) is active:
- If DBUZ is active, the utility obtains authorization through DBUZ.
- If DBUZ is not active, the utility proceeds to step 4.
The utility searches for an active Cross-Product Connectivity (CPC) subsystem:
- If a CPC subsystem is active in the environment, the utility obtains authorization through that subsystem and reports the subsystem ID in an informational message.
- If no CPC subsystem is active, the utility issues an error message and terminates abnormally.

The CPC subsystem is found automatically, you do not need to provide the subsystem ID.

HALDB Considerations for CAP

The Change Accumulation utility provides complete support for High Availability Large Databases (HALDBs).

The default value for the HALDB keyword is Y. You must not specify the HALDB(N) keyword on the GLBL statement in the ICPSYSIN data set of the utility job step. Additional considerations and instructions apply to use of the IIC function with HALDBs.

HALDB Considerations for ICP

The Image Copy utility provides complete support for High Availability Large Databases (HALDBs).

The following considerations apply to use of the Image Copy utility with HALDBs:

You must not specify the HALDB(N) keyword on the GLBL statement in the ICPSYSIN data set of the utility job step. The HALDB(Y) keyword allows the utility to establish the correct environment.
To process all data set groups in all partitions in the HALDB implicitly, specify the master DBD name (with the DBD keyword), omit the DDN keyword and specify the DBALLOC(Y) keyword on the GLBL control statement.
To process all data set groups within a partition implicitly, specify the master DBD name with the DBD keyword, and specify the partition name with the DDN keyword.
To process individual data set groups within a partition, specify the master DBD name with the DBD keyword and specify the data set group name with the DDN keyword.
The utility cannot create a virtual image copy, as specified with the VIC(Y) keyword, for the indirect list data set (ILDS) or the primary index of a HALDB.

HALDB Considerations for RVP

The Recovery utility provides support for High Availability Large Databases (HALDBs).

The following considerations apply to use of the Recovery utility with HALDBs:

You must not specify the HALDB(N) keyword on the GLBL statement in the RVPSYSIN data set of the utility job step.
To process all data set groups in all partitions in the HALDB implicitly, specify the master DBD name (with the DBD keyword), omit the DDN keyword, and specify the DBALLOC(Y) keyword on the GLBL control statement.
To process all data set groups within a partition implicitly, specify the master DBD name with the DBD keyword, and specify the partition name with the DDN keyword.
To process individual data set groups within a partition, specify the master DBD name with the DBD keyword and specify the data set group name with the DDN keyword.
If you specify the BLDILDS(N) keyword (meaning that you do not want to rebuild the ILDS) for a timestamp recovery, you must run the DFSPREC0 utility, even if no indirect list entries (ILEs) are present, to turn off the RECOVERY NEEDED flag for the ILDS component.
If you specify the BLDINDEX(N) keyword (meaning that you do not want to rebuild the primary indexes) for a timestamp recovery of a PHIDAM database, you must run the DFSPREC0 utility to turn off the RECOVERY NEEDED flag for the primary index component.
During a roll-forward recovery with the Recovery utility, primary indexes for PHIDAM databases cannot be rebuilt, as specified with the BLDINDEX(Y) keyword, or recovered, as specified with the REC command. ILDSs cannot be rebuilt during a roll-forward recovery. To recover primary indexes and ILDSs, use the Index Build function or use the HALDB Index/ILDS Rebuild (DFSPREC0) utility.

Point-in-time change accumulation considerations

Several considerations apply to the production of a point in time (PIT) change accumulation (CA). For details, see PIT change accumulation.

IMS HALDB Online Reorganization for CAP

The Change Accumulation utility provides toleration support for IMS HALDB Online Reorganization (OLR) as follows (if you have IMS 8.1 or earlier, OLR support is disabled):

A point-in-time (PIT) change accumulation cannot be performed to a time when OLR is actively performing a reorganization. No additional considerations apply to standard change accumulation processing.
The image copy process can be performed only when OLR is not actively performing a reorganization. The utility ignores any commands to create image copies if an online reorganization is in progress. Dynamic allocation for HALDBs is recommended so that the utility automatically can select the active data set group for processing. No control statement changes are necessary.
Incremental image copy processing is not allowed unless an input image copy is present.

IMS HALDB Online Reorganization for ICP

The Image Copy utility provides toleration support for IMS HALDB Online Reorganization (OLR) as follows (if you have IMS 8.1 or earlier, OLR support is disabled):

The image copy process can be performed only when OLR is not actively performing a reorganization. The utility ignores any commands to create image copies if an online reorganization is in progress.
BMC recommends dynamic allocation for HALDBs so that the utility automatically can select the active data set group for processing. No control statement changes are necessary.
Incremental image copy processing is not allowed unless an input image copy is present.
BMC recommends that you take an image copy (batch or concurrent) after an online reorganization.

IMS HALDB Online Reorganization for RVP

The Recovery utility provides toleration support for IMS HALDB Online Reorganization (OLR) as follows (if you have IMS 8.1 or earlier, OLR support is disabled):

Recovery to current is allowed. If OLR is active, the utility recovers both paired data set group sets (A-J and M-V).
Timestamp recovery is allowed to any time after the last OLR if the partition has been stopped with the /DBR command.
Timestamp recovery is allowed to a time during the last OLR if the OLR has been suspended and the partition has been stopped with the /DBR command.
Point-in-time (PIT) recovery is allowed to any time after the last OLR.
PIT recovery is allowed to a time during the last OLR if the OLR has been suspended and the partition has been stopped with the /DBR command.
PIT recovery is allowed if an OLR is active and if all data set groups are unallocated at the PIT recovery time. However, if one or more of the data set groups is allocated at the PIT recovery time, the recovery job is terminated abnormally. The failing data set groups are indicated in a SYSPRINT message.
The image copy process can be performed only when OLR is not actively performing a reorganization. The utility ignores any commands to create image copies if an online reorganization is in progress. BMC recommends dynamic allocation for HALDBs so that the utility automatically can select the active data set group for processing. No control statement changes are necessary. BMC recommends that you take an image copy (batch or concurrent) after an online reorganization.

DFSUCUM0 functions

The following control statement functions provided by DFSUCUM0 are not supported in CHANGE ACCUMULATION PLUS:

DB1 *ALL
DB1dbdname
SO

Recovery Manager (RMGR)

Recovery Manager (RMGR) functions and utilities are available with BMC AMI Backup and Recovery for IMS. The tasks that you can perform with RMGR include the following:

Create, validate, and rebuild groups of databases
Check recovery assets
Create recovery points
Synchronize log switches
Find recovery points
Analyze logs
Create recovery JCL
Maintain information about additional recovery assets
Issue selected IMS commands with a positive response
Automatically delete and define database data sets
Clean up RECON data sets in a disaster recovery situation

Recovery Advisor exceptions in the console

The Image Copy utility, the Change Accumulation utility, and the Recovery utility interact with Recovery Advisor in the console.

One important interaction is the automatic clearing of certain Recovery Advisor exceptions at successful creation of an image copy or change accumulation data set and at successful recovery of a database.

Supporting IMS-managed ACB

BRI utilities support the use of IMS-managed ACB. This functionality can be activated by:

Setting the BMC global option keyword IMSMACB to YES in the DLIGSET0, or by providing a BMCSETUP DD statement in the JCL containing IMSMACB(Y) keyword as follows:
//BMCSETUP DD *
IMSMACB(Y)
The default value for the IMSMACB keyword in DLIGSET0 is changing to Yes. This makes the job step eligible for the IMS-managed ACB processing by the BRI utilities.
Important
The default value has been "N" up through RSL2307.
Providing one of the following for the BRI utilities to verify that an IMS-Managed ACB environment exists:
- IMS Catalog user exit (DFS3CDX0) in the STEPLIB, JOBLIB or linklist library.
- DFSDFxxx Procedure library member. The required member suffix for the PROCLIB can be specified using the keyword PARM=’DFSDF=xxx’ on the job step’s EXEC statement and by including a //PROCLIB DD containing the matching parameter member.
Providing information for the utilities to find the IMS catalog data set high level qualifier by one of the following methods:
- An MDA member mapping the Catalog bootstrap data set.
- From the Catalog database definitions in the RECON datasets.

The utilities allow the IMS catalog ALIAS names defined using the catalog exit (DFS3CDX0) or the DFSDFxxx PROCLIB member to support the IMS environment with multiple catalogs.

The following table describes how the staging or active directory is accessed in the CHANGE ACCUMULATION PLUS, IMAGE COPY PLUS, and RECOVERY PLUS for IMS components.

Utility	ACBLIB environment	Directory to read the database		Alternate via keyword(NEWDBD- Value S)
Utility	ACBLIB environment	Default Directory	Alternate via keyword(NEWDBD- Value A)	Alternate via keyword(NEWDBD- Value S)
ICPUMAIN	IMSACB	ACTIVE	ACTIVE only	Search the STAGING library first, if not found then search in the ACTIVE directory.
RVPUMAIN	IMSACB	ACTIVE	ACTIVE only	Search the STAGING library first, if not found then search in the ACTIVE directory.
CAPUMAIN	IMSACB	ACTIVE	ACTIVE only	Search the STAGING library first, if not found then search in the ACTIVE directory.

Unsupported versions of IMS

If you are using a version of IMS that is no longer supported by IBM, new and recent enhancements to the BMC AMI Backup and Recovery for IMS product might be unavailable.

For example, dynamic allocation models are not available for IMS Version 6.1 or earlier.

General considerations for CAP

The following general considerations apply to the use of the Change Accumulation utility:

Change accumulation of a shared secondary index data set is supported if you specify the first shared secondary index in the change accumulation group.
You must use Change Accumulation utility format control statements (in the CAPSYSIN data set) if you process multiple change accumulation groups in a single job step.
The Change Accumulation utility does not support dynamic allocation of predefined image copy data sets for databases that are defined with the REUSE option in the DBRC RECON data set. If you use the REUSE option, you must allocate the correct image copy data sets in the JCL.
The Change Accumulation utility supports the IBM IMS Partition Support Product (Partition DB). However, the partitioned indexes that are available with the Partition DB version 2 are not supported.

General considerations for ICP

The following general considerations apply to the Image Copy utility:

It does not recognize IBM IMS Online Image Copy utility JCL.
It does not support the GSAM, HSAM, or MSDB database organizations.
It does not support the IBM Utility Control Facility (UCF).
It cannot process the same data set group or area more than once in the same job step. This limitation includes shared secondary indexes.
The Online Image Copy function requires the data set to be registered with DBRC.
The Incremental Image Copy function and the ICP Database Recovery function do not process log tapes.
The ICP Database Recovery function does not support the following:
- DEDBs
- roll-forward recovery operations (an input image copy is required)
- log input
- dynamic allocation of input image copy or change accumulation data sets for recovery processing
Asynchronous processing of HISAM databases is not supported if you also request pointer checking with the PCP(Y) keyword. If you specify PCP(Y) and have HISAM databases with overflow data sets, the Image Copy utility automatically changes asynchronous processing requests to synchronous processing requests.
The Image Copy utility supports creating an image copy of a shared secondary index data set if, on the IC control statement, you specify the first shared secondary index by using the DBD keyword. The Virtual Image Copy feature does not support shared secondary indexes.
The Image Copy utility does not support dynamic allocation of predefined image copy data sets for databases that are defined with the REUSE option in the DBRC RECON data set. If you use the REUSE option, you must allocate the correct image copy data sets in the JCL.

IMS Partition DB Support

The Recovery utility supports the IBM IMS Partition Support product (Partition DB).

All partitions must be available (through dynamic allocation or explicit allocation in the JCL) to be scanned if you want the Recovery utility to rebuild the indexes. The Recovery utility supports the rebuilding of partitioned indexes as long as the default value HALDB(Y) is in effect.

KSDS support

Support for VSAM KSDS databases is limited as follows:

The Image Copy by Volume feature works only with an OSAM or VSAM ESDS. It does not support VSAM KSDS.
The ICP Database Recovery function does not support track recovery or recovery of only the index portion of a VSAM KSDS.
The Online Image Copy function works with a VSAM KSDS, but you must take certain steps to ensure data integrity.
The Snapshot Copy function does not support a VSAM KSDS that was defined with the IMBED and REPLICATE parameters.
The Instant Snapshot function does not support a VSAM KSDS that was defined with KEYRANGES.

Snapshot Copy function in ICP

The Snapshot Copy function requires the use of SUF.

The following restrictions apply to the Snapshot Copy function:

The Snapshot Copy function does not support a VSAM key sequenced data set (KSDS) that was defined with the IMBED and REPLICATE parameters.
VSAM data sets must use integrated catalog facility (ICF) catalogs.
For VSAM data sets, the VSAM SHAREOPTIONS parameter must be set to (2,3) or greater. IMS requires that the VSAM SHAREOPTIONS parameter be set to (3,3) or greater if databases are registered to DBRC at SHARELVL 2 or 3, or SHARELVL 1 when you are using IRLM.
The Snapshot Copy function does not support VSO DEDBs. Attempts to include a VSO DEDB in a Snapshot Copy job result in a SUF registration failure, as indicated by the following message:
BMC8300C SNAPSHOT COPY REGISTRATION FAILED
BMC8300C DATA SET IS NOT CACHEABLE BY XBM. REASON CODE=1080
The Snapshot Copy function does not support block or CI sizes that are less than 512 bytes. An attempt to perform the Snapshot Copy function for a database that does not meet this restriction causes the function to issue the following message:
BMC8300C DATA SET IS NOT CACHEABLE BY XBM. REASON CODE=1080
If this error occurs, the Snapshot Copy function will fail for any other registered data set. The function will issue message BMC73553I with return code 12 and reason code 1098.

Instant Snapshot function in ICP

The Instant Snapshot function requires the use of SNAPSHOT UPGRADE FEATURE.

The following requirements and limitations apply to the Instant Snapshot function:

The database to be copied must reside on a supported intelligent storage device.
The Instant Snapshot data set name cannot be a GDG.
The database to be copied must be registered with DBRC.
You must use dynamic allocation for output image copy data sets.
Because the logical database structure is not read during Instant Snapshot processing, pointer checking is not available.
The Instant Snapshot function does not support KSDS databases that are defined with KEYRANGES.

BMC AMI Application Restart Control for IMS

If you want to use the optional BMC AMI Application Restart Control (AR/CTL) suspend-and-resume interface for automatic handling of BMPs during the Snapshot Copy process, you must have BMC AMI Application Restart Control for IMS, version 2.2.00 (or later).

If you have installed AR/CTL from a tape with a VOLSER earlier than AES94x, contact BMC Customer Support for replacement load modules.

Track recovery

The Recovery utility does not support track recovery.

HISAM UNLOAD

The Recovery utility does not support the use of HISAM unload output for the ESDS portion of a database.

HISAM unload output can be used for the KSDS portion. You can also use HISAM unload on a primary or secondary index.

HISAM database processing

The Recovery utility does not support asynchronous processing of HISAM databases if you are performing concurrent pointer checking, as requested with PCP(Y).

If you specify PCP(Y) and have HISAM databases with overflow data sets, the Recovery utility automatically changes an asynchronous recovery (AREC) to a synchronous recovery (REC).

IMS Utility Control Facility (UCF)

The Recovery utility does not support the UCF.

Shared secondary indexes

The Recovery utility supports recovery of a shared secondary index if, on the REC control statement, you specify the first shared secondary index using the DBD keyword.

The Recovery utility does not support the Index Rebuild function for rebuilding shared secondary indexes.

Data Base Image Copy 2 data sets

The Recovery utility supports the use of data sets created by the Data Base Image Copy 2 utility from IBM (program DFSUDMT0, also referred to as the T0 Copy utility).

When you use a T0 copy as input, the following considerations apply:

The utility uses the two-phase recovery method to restore the database.
Asynchronous recovery requests (with AREC control statements) are processed synchronously (as if requested with REC control statements).
For recovery of a Fast Path multiple area data set (MADS), the utility restores only the first area data set. The utility ignores the rest and issues a warning message.
All DFSMSdss messages are routed to the ADRPRINT data set. This data set is dynamically allocated to SYSOUT if you do not provide it in the JCL. ADRPRINT cannot be allocated as a dummy data set.