Configuration Best Practices › Enable the History Database for Spill Statistics

Enable the History Database for Spill Statistics

The History Database table CADTCM_HISTORY_SPL provides a persistent record of DBUTLTY SPILL executions. Detailed information is provided about each of the spills of the Log Area (LXX) to the Recovery File (RXX). This information can be used to determine the contents of each recovery file.

This information can be helpful when you find that you have a damaged database and need to recover the database from a previous backup. Once the backup is reloaded to the database, the database is at the point in time of the backup.

The spill history records help you to determine what RXX files are necessary to run a DBUTLTY forward recovery to bring the restored database forward in time.

Business Value

Typically when database damage occurs, and the user must restore a backup and do forward recovery, time-to-recover is of great importance. Unfortunately, the need to recover quickly can add additional pressure to the job of creating the recovery JCL.

If only one of the RXX files is missed or if the RXX files are specified in the wrong order, the forward recovery can fail and cause the user to restore the backup and run the recovery a second time.

By having the CADTCM_HISTORY_SPL table available, you can quickly determine all of the RXX files that have been created since the backup in question was created. Verify that these RXX files are provided in the correct sequence to the forward recovery process. This information not only speeds the job of creating the recovery JCL, but also improves the accuracy of the JCL creation.

Additional Considerations

The History Database automatically contains the additional CADTCM_HISTORY_SPL table in any system that has been installed or upgraded to the Version 12.0 level or later. This new table shares the same data area (A01) as the other History Database table CADTCM_HISTORY_ADS. Verify that the Index and data areas have been allocated, initialized and loaded null before starting the History Database.

The MUF startup option HISTORY must be added to the MUF startup to enable the statistical collection.

HISTORY dbid

dbid

Provides the database ID for the History Database. The standard implementation uses DBID 1007 for the History Database.

The spill statistics are created each time a DBUTLTY SPILL process is executed in a MUF environment with the LOGRCV=NO option specified.

For more information about using the History Database, see the CA Datacom/DB Database and System Administration Guide.

Using Shadow MUF

Use the Shadow MUF to extend the ability for 24x7 MUF availability.

Business Value

The Shadow MUF is a nearly identical MUF job or started task that is usually running on a different LPAR within the z/OS Sysplex. Functionally, the Shadow MUF is waiting in stand-by mode ready to take over the CA Datacom processing should the Primary MUF need to be brought down on goes down due to any unforeseen or unrecoverable error.

Additional Conditions

Typically, the CA Products that are using the AD repository recommend if Shadow MUF processing is needed for this product.

Hot Backups versus Cold Backups

Depending on your requirements, backups generally fall into one of two categories:

• Hot backup - performed when the database or DBMS is still up and servicing read and maintenance requests
• Cold backup - taken when the database or DBMS is down

Business Value

Performing a hot backup provides the following advantages:

• Database and DBMS are actively available to the applications
• DBULTY BACKUP SEQ=PHYSOCAL and non-DBUTLTY backups can be taken at any time.

Performing a cold backup offers the following advantages:

• Simplest form of backup taken whenever all MUFs that reference the database are down.
• Absolute guarantee that all data is backed up
• DBUTLTY BACKUP SEQ=NATIVE can be taken in preparation of a corresponding LOAD which ultimately reorganizes the data into native key sequence. This accesses the records by reading the index for the lowest native key value and retrieving the corresponding data record. This requires exclusive access to the database.
• DBUTLTY BACKUP SEQ=PHYSICAL is the fastest of DBUTLTY backups because it does not read the index at all. It simply starts from the beginning of the data area and copies each record it finds to the backup file. This does not require exclusive access to the data area.

Enlarging the LXX Log File

Enlarging the CA Datacom LXX allows continuous processing.

Business Value

The LXX that is allocated as part of a normal CA Datacom install is sized for an average amount of data records being updated. Depending on the number of records that are updated in some fashion by either an application or the CA Datacom DBMS and how often you do or do not want to spill, we recommend that you enlarge the size of your LXX.

Finding Compatibility Information on CA Datacom

To find the operating systems or CICS release compatibility for any of the CA Datacom family of products go to our support web site at http://support.ca.com. After signing in, click on COMPATIBILITIES in the left column of the support home page. From the CA Product Compatibilities page, select the platform of interest then select the DATABASES category from the platform compatibility page. On the DATABASES category page, find the product, platform release and product release level. The informational solution number that contains the desired information is located on the far right hand side.

Business Value

This process enables you to obtain faster resolution or queries of your research and fewer support calls.