Wait type PAGEIOLATCH_SH

  • Being new to performance tuning in SQL Server, I am curious as to the PAGEIOLATCH_SH, is there a way to clear these with out having to reboot the server.

    As you can see by the data below, the drop in numbers relates to a reboot of the server. We do have 6 remote locations that are on VPN at this time, I also have 8 other remotes on MPSL.

    Any help or knowledge I can gain will improve my life.

     

    'dt''wait_type''waiting_tasks_count''wait_time_ms''max_wait_time_ms''signal_wait_time_ms'
    7/5/2007 0:00PAGEIOLATCH_SH5772784943515245321921
    7/5/2007 2:00PAGEIOLATCH_SH5782624946015245321921
    7/5/2007 4:00PAGEIOLATCH_SH5787504948437245321937
    7/5/2007 6:00PAGEIOLATCH_SH5790974950406245321937
    7/5/2007 8:00PAGEIOLATCH_SH5886035063703245323015
    7/5/2007 10:00PAGEIOLATCH_SH52184364051562
    7/5/2007 12:00PAGEIOLATCH_SH786274796546140
    7/5/2007 14:00PAGEIOLATCH_SH10004107640656171
    7/5/2007 16:00PAGEIOLATCH_SH10898117281656171
    7/5/2007 18:00PAGEIOLATCH_SH12428126312656171
    7/5/2007 20:00PAGEIOLATCH_SH12663129921812187
    7/5/2007 22:00PAGEIOLATCH_SH12692133140812187

     

    Thanks

    Greg

  • Well I just gained more knowledge on my own, sys.dm_os_wait_stats, the stats keep adding up until you run the following:

    DBCC

    SQLPERF ('sys.dm_os_wait_stats', CLEAR);

    Now my monitoring will gather stats stored every two hours so that I will get a better representation of the true statistics.

    I love it when I teach myself

  • PAGEIOLATCH_SH simply means that the SPID is waiting for some page to be brought into memory (from database files on disk).

    Investigate if your IO is a bottleneck. or see if you can tune your queries to reduce IO.


    -- Amit



    "There is no 'patch' for stupidity."



    Download the Updated SQL Server 2005 Books Online.

  • If you want a good idea of how many latch waits you're getting, use Performance Monitor. The counters you want are Latches:Average latch wait time, Latches: Latch waits/sec

    The average wait time should be low. Latches are supposed to be light-weight, short duration synchronisation devices. If the latch waits are high, monitor your physical disk stats. sec/read and sec/write

    Your sec/write should be under 10 ms

    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 4 posts - 1 through 3 (of 3 total)

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