none
TCP error 26073; RRS feed

  • Question

  • On sql 2008 Enterprise edition sp3 cu2(10.0.5768) seeing this error,(win2k8r2 sp1).

    TCP connection closed but a child process of SQL Server may be holding a duplicate of the connection's socket.  Consider enabling the TcpAbortiveClose SQL Server registry setting and restarting SQL Server. If the problem persists, contact Technical Support.

    But already have all the updates mentioned here http://support.microsoft.com/kb/2491214

    Any reason we still get this error?
    Thanks


    • Edited by SQL_Jay Monday, April 23, 2012 8:35 PM link
    Monday, April 23, 2012 8:35 PM

Answers

  • Hi SQL _Jay,

    Regarding to your description and error message you provided, which seems this issue occurs because SQL Server Native Client 10.0 handles a TCP close call on a socket incorrectly as you mentioned.

    If you tired all workarounds in the KB as you provided but still have problem, please post the SQL Server 2008 error log file.

    Alternatively you can try to install latest cumulative update: Cumulative update package 3 for SQL Server 2008 Service Pack 3. You can get from here: http://support.microsoft.com/kb/2648098/en-us

    Regards, Amber zhang

    • Marked as answer by SQL_Jay Tuesday, April 24, 2012 7:15 PM
    Tuesday, April 24, 2012 8:53 AM
    Moderator

All replies

  • Hi SQL _Jay,

    Regarding to your description and error message you provided, which seems this issue occurs because SQL Server Native Client 10.0 handles a TCP close call on a socket incorrectly as you mentioned.

    If you tired all workarounds in the KB as you provided but still have problem, please post the SQL Server 2008 error log file.

    Alternatively you can try to install latest cumulative update: Cumulative update package 3 for SQL Server 2008 Service Pack 3. You can get from here: http://support.microsoft.com/kb/2648098/en-us

    Regards, Amber zhang

    • Marked as answer by SQL_Jay Tuesday, April 24, 2012 7:15 PM
    Tuesday, April 24, 2012 8:53 AM
    Moderator
  • This issue occurs because SQL Server Native Client 10.0 handles a TCP close call on a socket incorrectly
     
    When an application connects to SQL Server 2008 or SQL Server 2008 R2 by using Microsoft SQL Server Native Client 10.0, you receive a "Non-yielding Scheduler" error intermittently.

    Options
     
    1) Install latest cumulative update
     
    For SQL Server 2008 R2
     
    http://support.microsoft.com/kb/981356  - The SQL Server 2008 R2 builds that were released after SQL Server 2008 R2 was released
     
     For SQL Server 2008 Service Pack 2
     
    http://support.microsoft.com/kb/2402659/LN  The SQL Server 2008 builds that were released after SQL Server 2008 Service Pack 2 was released
     
    2) For SQL Server 2008 R2
     
     The fix for this problem was introduced at cumulative update 6 - http://support.microsoft.com/kb/2489376
     
    For SQL Server 2008 Service Pack 2 - The fix for this problem was introduced at cumulative update 6 -


    Please click "Propose As Answer" if a post solves your problem. OR "Vote As Helpful" if a post has been useful to you. Thanks, Debasish Mohanty Blog:http://sqlcodebank.blogspot.com

    Tuesday, April 24, 2012 1:35 PM
  • I still experience this issue on SQL Server 2008 Service Pack 3.
    Tuesday, April 24, 2012 6:27 PM
  • Yes Like Jon Morisi mentioned our sql server already has SP3 installed, build is 10.0.5768. Also OS has sp1 installed (win2k8 R2).

    These were the events when we see this error

    Thanks

    2012-04-22 21:03:22.36 Server      Error: 26073, Severity: 16, State: 1.
    2012-04-22 21:03:22.36 Server      TCP connection closed but a child process of SQL Server may be holding a duplicate of the connection's socket.  Consider enabling the TcpAbortiveClose SQL Server registry setting and restarting SQL Server. If the problem persists, contact Technical Support.
    2012-04-22 21:03:32.85 Server      Error: 26073, Severity: 16, State: 1.
    2012-04-22 21:03:32.85 Server      TCP connection closed but a child process of SQL Server may be holding a duplicate of the connection's socket.  Consider enabling the TcpAbortiveClose SQL Server registry setting and restarting SQL Server. If the problem persists, contact Technical Support.
    2012-04-22 21:03:32.87 spid91s     Error: 1479, Severity: 16, State: 1.
    2012-04-22 21:03:32.87 spid91s     The mirroring connection to "TCP://xxxxxxxxxx" has timed out for database "xxxx" after 30 seconds without a response.  Check the service and network connections.
    2012-04-22 21:03:41.45 spid91s     Database mirroring is inactive for database 'xxxx'. This is an informational message only. No user action is required.
    2012-04-22 21:03:41.56 Server      Error: 26073, Severity: 16, State: 1.
    2012-04-22 21:03:41.56 Server      TCP connection closed but a child process of SQL Server may be holding a duplicate of the connection's socket.  Consider enabling the TcpAbortiveClose SQL Server registry setting and restarting SQL Server. If the problem persists, contact Technical Support.
    2012-04-22 21:03:43.45 spid614     Error: 18056, Severity: 20, State: 29.
    2012-04-22 21:03:43.45 spid614     The client was unable to reuse a session with SPID 614, which had been reset for connection pooling. The failure ID is 29. This error may have been caused by an earlier operation failing. Check the error logs for failed operations immediately before this error message.
    2012-04-22 21:03:46.02 spid621     Error: 18056, Severity: 20, State: 29.
    2012-04-22 21:03:46.02 spid621     The client was unable to reuse a session with SPID 621, which had been reset for connection pooling. The failure ID is 29. This error may have been caused by an earlier operation failing. Check the error logs for failed operations immediately before this error message.
    2012-04-22 21:03:46.02 spid528     Error: 18056, Severity: 20, State: 29.
    2012-04-22 21:03:46.02 spid528     The client was unable to reuse a session with SPID 528, which had been reset for connection pooling. The failure ID is 29. This error may have been caused by an earlier operation failing. Check the error logs for failed operations immediately before this error message.
    2012-04-22 21:03:46.02 spid351     Error: 18056, Severity: 20, State: 29.
    2012-04-22 21:03:46.02 spid351     The client was unable to reuse a session with SPID 351, which had been reset for connection pooling. The failure ID is 29. This error may have been caused by an earlier operation failing. Check the error logs for failed operations immediately before this error message.
    2012-04-22 21:03:46.02 spid355     Error: 18056, Severity: 20, State: 29.
    2012-04-22 21:03:46.02 spid355     The client was unable to reuse a session with SPID 355, which had been reset for connection pooling. The failure ID is 29. This error may have been caused by an earlier operation failing. Check the error logs for failed operations immediately before this error message.
    2012-04-22 21:03:46.02 spid618     Error: 18056, Severity: 20, State: 29.
    2012-04-22 21:03:46.02 spid618     The client was unable to reuse a session with SPID 618, which had been reset for connection pooling. The failure ID is 29. This error may have been caused by an earlier operation failing. Check the error logs for failed operations immediately before this error message..............

    Tuesday, April 24, 2012 6:58 PM
  • We are seeing this error also with SQL Server 2008 SP3 10.0.5500, running on a cluster with Windows Server 2008 SP1 x64.

    We see similar pattern of error 26073, followed by error 18056. The first error in the sequence, for us, is an 'I/O request taking longer than 15 seconds to complete...' error.

    It happens for a 1 hour period, about once a day, on the weekend, then stops happening.

    SQL_Jay - did you solve it? Please contact me directly - jason_daly@trimble.com.

    Microsoft - wtf? This isn't cheap software, please post an answer.

    Regards,

    Jason

    Sunday, August 5, 2012 10:41 PM
  • Do you use Sophos?
    Monday, August 6, 2012 2:00 AM
  • We are seeing this error also with SQL Server 2008 SP3 10.0.5500, running on a cluster with Windows Server 2008 SP1 x64.

    We see similar pattern of error 26073, followed by error 18056. The first error in the sequence, for us, is an 'I/O request taking longer than 15 seconds to complete...' error.

    It happens for a 1 hour period, about once a day, on the weekend, then stops happening.

    SQL_Jay - did you solve it? Please contact me directly - jason_daly@trimble.com.

    Microsoft - wtf? This isn't cheap software, please post an answer.

    Regards,

    Jason

    Yes we don't see the issue anymore did you apply this fix?http://support.microsoft.com/kb/2491214
    F
    or the IO request taking longer than 15 seconds you might have to work with your storage team.

    Tuesday, August 7, 2012 10:49 AM
  • the issue is not resolved after applying SP2 CU7 on SQL Server 2008 R2... 

    Will the workaround suggested in http://support.microsoft.com/kb/2491214 work? 

    TCP connection closed but a child process of SQL Server may be holding a duplicate of the connection's socket.  Consider enabling the TcpAbortiveClose SQL Server registry setting and restarting SQL Server. If the problem persists, contact Technical Support.


    Suman

    Wednesday, October 23, 2013 10:26 PM
  • hi Jason, 

    Is this problem closed? can u please update what you did to get over it... 


    Suman

    Wednesday, October 23, 2013 10:27 PM
  • hi Jason, 

    Is this problem closed? can u please update what you did to get over it... 


    Suman

    Hello Everyone,

    I experienced same issue. We have SQL 2008 R2 SP3 (build 10.50.6000.34) on windows server 2008 R2 enterprises SP1.

    what is the solution for this.

    Thanks

    Sunday, January 10, 2016 5:46 PM