IBM Db2 Analytics Accelerator for z/OS Version 7.1.6 Release Notes

General recommendations

• Do not run more than 30 load threads in parallel.
• Do not use partial reload for replicated tables.
• Wait 30 seconds before you route queries to a newly paired accelerator.
• Do not use interactive load recommendations and batch job change detection at the same time (to ensure you receive correct reload recommendations)
• When reloading multiple incrementally updated tables, please pause after each reload, as this allows the system to come back to normal throughput rates and latency levels.
• Reduce the task diversity of parallel jobs if you process big table sets with more than 100 tables. For example, do not run 'Alter Keys', 'Add Tables', and 'Remove Tables' operations at the same time. Reduce this to 'Alter Keys' and 'Add Tables' jobs or any other combination of just two different tasks.
  If a query failed, rerun the query.
• For accelerator-only tables do not use the multi-row insert method for rows containing a NULL value for columns with data type BINARY or VARBINARY.
• Before using SYSPROC.ACCEL_ALTER_TABLES with partitioned tables that are replicated, issue a SYSPROC.ACCEL_LOAD_TABLES to avoid lost rows on the accelerator.

Dealing with unexpected behavior of the accelerator

• Retry the operation when it fails (potentially with an error message), e.g. retry the load operation or restart the Incremental Update process or restart aborted tasks.
• When Add-tables-jobs fail, and no clean-up job is started thereafter, please manually remove the table from the accelerator and then add it again.
• Disable replication for tables and then re-enable replication if bulk operation failed or tables are in invalid state.
• Reload tables that are no longer replicated although they are enabled for replication.

• Data on an accelerator might be uncompressed or the compression rate is lower than the rate used by Db2 for z/OS.
• Some table-level statistics are displayed incorrectly.
  • The table size shown for each table in the accelerator view of the IBM Db2 Analytics Accelerator Studio is too large.
  • The table-level statistics displayed in the IBM Db2 Analytics Accelerator Studio is incorrect. Table sizes statistics do not reflect the actual usage on the accelerator. For example, the table size metrics do not reduce when data is deleted from the table. However, the total space usage on the accelerator (Q8STDSA / Q8STDSKB) is correct.
  • The value for the monitoring counter Q8STTMUD are to high, e.g. 0.18E+20. The high value can be ignored. It will be corrected during the next statistics collection.
  • The values for monitoring metrics Q8STCQL, Q8STCQLS, and Q8STQUEW may be displayed incorrectly.
• In Db2z installation which are configured for DBCS the stored procedure SYSPROC.ACCEL_GET_TABLES_DETAILS returns an error.

Recommendations specific to IBM Db2 Analytics Accelerator on IBM Z

• Use only FICON DASD or only FCP storage devices for a specific storage pool definition since the accelerator does not support a combination of FICON DASD and FCP storage devices within the same storage pool.
• If the query performance with large memory configuration is low please contact IBM to get support with memory configuration and reload recommendation.
• Avoid using an 'EXCEPT ALL' clause in queries executed on the accelerator.