Background documentationPhysical Reads for User Task Locate this document in the navigation structure

 

<number of> physical reads for user task <task_ID>, <number of> commands, application pid <pid>

Explanation

The user task <task_id> causes a large number of physical reads to take place on the volumes of the database, because the data requested by the applications was not found in the data cache.

User Response

If a table is accessed for the first time, or if it has not been used for a long time, and was therefore displaced from the data cache, then this situation is not a problem.

However, if this does not explain the read activity, you should check the hit rate for the data cache, and increase the size of the data cache, if necessary.

Make sure that the SQL statements specified by the application do not read significantly more data than is necessary for processing, because of poor search strategies, for example.

If the special database parameter UseDataCacheScanOptimization has the value yes, then table scans (TABLE SCAN) use only 10% of the data cache for caching the table. This means that it may not be possible to hold the entire table in the data cache, and the next scan has to read it physically again.