Previous Topic: Limit Harvested SQL

Next Topic: Use Summarization for Harvested Data

Data Collection Granularity Qualifiers

During the monitor process, data is collected in an effort to pinpoint opportunities to improve performance.

In DB2, the object of interest is the SQL statement and the name of the package that it belongs to. Sometimes knowing what package the SQL statement belongs to is sufficient information in a standard batch type job, but it is not enough information in online or distributed environments. For example, in CICS, either the transaction name or the DB2 user ID might be useful.

The Synchronous Data Gatherer provides four additional data qualifies to assist in analyzing monitored SQL statements. Some of these qualifiers are automatically turned on, depending on the type of DB2 connection.

CA Mainframe Application Tuner provides these commands in the TUNSSP00 member.

DB2HVLOC

Harvest the requesting location.

The requesting location name is generally of importance for distributed DB2 threads. It can help to identify where the SQL originated from.

Most other threads will originate from the local location.

The following are valid responses:

YES - Use LOCATION as part of the key to gather information.

NO - Do not add LOCATION to the key.

The default is DB2HVLOC=YES.

DB2HVCOR

Use correlation ID as part of the key.

Collect the DB2 correlation ID for the DB2 data in both the Harvester and the sampling process.

This option adds a higher level of granularity to the DB2 data being collected.

For IMS, the correlation ID is the PSB name.

For CICS, the correlation ID is the transaction name.

For Batch type threads, the correlation ID is the job name.

For remote and RRSAF threads, the value in this field is dependent on the requesting application.

The following are valid responses:

YES - Use the DB2 correlation ID in building the key.

NO - Do not add correlation ID to the key.

The default is DB2HVCOR=YES.

DB2HVIID

Use operator ID as part of the key.

Collect the DB2 primary operator ID for the DB2 data in both the Harvester and the sampling process.

This option adds a higher level of granularity to the DB2 data being collected.

The following are valid responses:

YES - Use the DB2 operator ID in building the key.

NO - Do not add operator ID to the key.

The default is DB2HVIID=YES.

DB2HVTHD

Use thread address as part of the key

Add the DB2 Thread Agent Control Element address to the data being collected by the Harvester and the sampling process.

The following are valid responses:

YES - Use the DB2 thread address in building the key.

NO - Do not add the DB2 thread address to the key.

The default is DB2HVTHD=YES.