Results 1 to 4 of 4
  1. #1
    Join Date
    May 2012
    Canberra, Australia
    Provided Answers: 6

    Unanswered: Db2 10.5 pool_read_time on insert

    DB2 10.5 FP1 on Linux
    We are running an activities event monitor in an attempt to determine causes of long running sql. One of the outcomes is that we can see significant pool_read_time on an insert statement. The definition of the pool_read_time monitor element is "Indicates the total amount of time spent reading in data and index pages from the table space containers (physical) for all types of table spaces. This value is given in milliseconds"
    I am trying to understand the events relating to the insert in regard to bufferpool activity and the reason for the pool_read_time. I expected that the insert would be writing to bufferpool data pages with free space where possible but also that it would have to alter existing index pages so may be reading them in from disk where necessary. Is this a reasonable conclusion? Or do I have a misperception of the background mechanics of and insert action?


  2. #2
    Join Date
    Apr 2012
    Provided Answers: 17
    Is the INSERT statement using a format like "Insert into ... select ... from ... where..." ?
    Get the full text of the SQL concerned and examine the access plan.

  3. #3
    Join Date
    Jun 2003
    Toronto, Canada
    Provided Answers: 1
    Some other things that can cause reads on insert are constraints and triggers on the table.
    "It does not work" is not a valid problem statement.

  4. #4
    Join Date
    Dec 2007
    Richmond, VA
    Provided Answers: 5
    lets not forget RI. Do you maybe not have an index on a foreign key?

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts