GA enhancements and changes in Db2 Query Monitor 3.4.0
The following enhancements are available as of the General Availability (GA) date of DB2® Query Monitor 3.4.0:
- Some of the Db2 Query Monitor processing done on the Db2 DBM1 address space is now moved to the Query Monitor address space to reduce Db2 class II (2) CPU time.
- Query Monitor now uses the AUTHID parameter to perform an AUTHID_SIGNON when installing its monitoring agent to check whether Db2 is ready. The user ID of the started task defined in the RACF STARTED class must be able to read and update all the files that the Query Monitor started task can access, such as back store, interval file, and profile data set.
- You can now stop the CAE Server job using a console
'
P
' command instead of issuing the command over the network. See CAE prerequisites.Note:The CAE Server requires the most current maintenance of 64-bit IBM Semeru Runtime Certified Edition Java 17.
- Db2 Query Monitor 3.4.0 removes the restriction where object filters are only applied in the structural drill-downs and the operational filters only in operational drill-downs. You can now apply filters at a much higher level in the data hierarchy.
- New fields are introduced in the Buffer Pool Statistics screen to access the metrics related to zHyperlink. See Buffer pool statistics - fields (ISPF).
Additionally, plan for the following changes when you migrate to version 3.4.0:
The DB2_CPU value, as reported in activity summaries and exceptions, now includes zIIP usage. Previously, in version 3.3.0, the DB2_CPU value included only central processor time.
- Because some of the Db2 Query Monitor processing is now moved to the Query Monitor address space, an increase in the CPU/zIIP time is expected after migrating to Db2 Query Monitor 3.4.0. For information on reducing the amount of zIIP used by the Query Monitor address space, refer to ZIIP_DURATION.
- The MAXIMUM_ALLOCATION and SMEM_SIZE values that worked for Db2 Query Monitor 3.3.0 might not work for Db2 Query Monitor 3.4.0 due to internal Query Monitor processing changes. Hence, these values require an increase when monitoring heavy workloads. Additionally, you will notice an increase in the shutdown time for the started task.