none
Wait stats

    Question

  • Can anyone provide more information regarding the PREEMPTIVE_OS_WAITFORSINGLEOBJECT wait type?

    I have a client with and Informatica session that BULK INSERTS a flat file from a Linux server to SQL Server 2008. One of the wait types with several hundred milli second waits is PREEMPTIVE_OS_WAITFORSINGLEOBJECT. I have a strong feeling it's related to the distributed nature of the insert.
    Friday, August 28, 2009 8:32 PM

Answers

All replies

  • Scalability Doug,

     As you may realized by this time these wait types aren’t documented yet. The only valuable reference I found is from Christian Bolton’s blog and guess you may have already tried that.

     

     As per him, these waits are outside the control of the SQL Server and more of the OS just as you thought.

    Just in case, if you haven't looked at, here is his post. http://sqlblogcasts.com/blogs/christian/archive/2008/06.aspx

    At some point of Jasson Massie (@statisticsio) seem to got these and try to ping him on twitter to get more details.

     

     

     


    | Sankar Reddy | http://sankarreddy.spaces.live.com/ |
    Saturday, August 29, 2009 3:20 AM
    Moderator
  • This wait type maps to a call to the Win32_WaitForSingleObject API and shows that SQL Server is waiting for the external call using preemptive scheduling to complete.  You can get information about the API from:

    http://msdn.microsoft.com/en-us/library/ms687032(VS.85).aspx

    A few hundred milliseconds isn't necessarily problemattic.  Can you provide more information about the process/problem that you are dealing with. 


    Jonathan Kehayias
    http://sqlblog.com/blogs/jonathan_kehayias/
    http://www.twitter.com/SQLSarg
    http://www.sqlclr.net/
    Please click the Mark as Answer button if a post solves your problem!
    Monday, August 31, 2009 3:27 AM
    Moderator
  • Thanks Jonathan

    We have been seeing a lot of both this and the PREEMPTIVE_OS_CRYPTOPS waits recently. We were able to track the CRYPTOPS down to a problem with linked servers on 2012 communicating with data held in 2008R2 (something to do with new security changes). With this; however, we have no idea what to do about it or how to prevent it in the future. MS should never had published these wait types sans documentation and needs to provide guidance as soon as possible so we don't have to call in for support.
    Monday, December 1, 2014 12:31 PM