Enterprise Edition
Customer Service

Brother-Eagle Community

  • Visit/Join Now!
  • Get tips from other users!
  • Share your own tips, advice, and metric formulas!

DBI Products

Brother-Eagle® DB2 Performance
Advice: Sync Read Pct

The synchronous read percentage indicates the proportion of read I/O that is completed with synchronous reads.

When DB2 has to scan many pages to find result sets, usually because indexes are missing or sub-optimally defined, DB2 uses prefetch I/O which is performed asynchronously. When the percentage of asynchronous I/O is high, then synchronous I/O is low.

When the Synchronous Read Percentage is low, DB2 is doing a lot of scanning to find result sets. Scans occur when indexes are missing or are sub-optimally defined.

OLTP Database Guidelines

When the Average Result Set Size is less than ten, an OLTP database is typically indicated.

For an OLTP database, the Synchronous Read Percentage should generally be 90% or higher. Brother-Eagle will indicate a warning if it is lower than 90%, and a severe alert condition if it is lower than 80%.

If the Synchronous Read Percentage is less than 80%, DO NOT UPGRADE YOUR HARDWARE. Instead, identify SQL that is the driving force behind high numbers of asynchronous reads (scans) and work to create or modify indexes to reduce scans. For optimum results in the shortest amount of time, DBI recommends Brother-Panther™ for DB2 LUW.

Data Warehouse Database Guidelines

When the Average Result Set Size is greater than, or equal to, ten, then a data warehouse database is typically indicated.

Data Warehouse databases tend to do a lot more data scanning than OLTP databases, but even these should make use of well defined indexes. Generally, for a data warehouse database, the Synchronous Read Percentage should be at least 25% or higher. If it is less than 10%, the need for indexes is strongly indicated. Look for high cost SQL with frequency of execution greater than one over a period of a few hours, and attempt to identify indexes, Multi-Dimensional Clustering Tables (MDCs), or Materialized Query Tables (MQTs) to improve your data warehouse performance.


View full list of DB2 LUW advice topics