Partitioned tables and statistics

  • Hi,

    Since a few months, we are working with partitioned tables.

    The tables are partitioned using a sliding window with 'Month' partitions for the last 5 years, and the last 2 months into 'Day' partitions.

    The Day partitions contains about 200000 rows, the whole table about 90000000.

    What do we see: after about 5 days, some queries become very slow. After executing the 'Update statistics', the performance is OK.

    Can anyone explain the reason of the bad performance after only 5 days?

    Does anyone know when it is needed, or how it is possible to define when it is needed, to execute the 'Update Statistics'?

    I know it is not possible to update the statistics of 1 partition. So the update statistics must be done of the whole table, and this is a costly operation.

    Thanks for any advice.

  • Coincidental, I wrote about what you're probably experiencing

    http://sqlinthewild.co.za/index.php/2011/03/22/statistics-row-estimations-and-the-ascending-date-column/

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass

Viewing 2 posts - 1 through 1 (of 1 total)

You must be logged in to reply to this topic. Login to reply