locked
SQL Server 2000 service stops RRS feed

  • Question

  • Thank you kindly for taking the time to read this, any response would be much appreciated.

    Problem:  After doing a P2V conversion of our SQL 2000 / Windows 2003 server the SQL service stops running right around midnight.

    Why does it do this? 

    I've checked the logs to find out that the memory was maxing out around 90+% so I've doubled the RAM on the machine (see here for dump file)

    I was sure that increasing the memory was the solution but it crashed again this morning.  (see here for dump file)

    Can anyone help me figure out why the service is stopping?

    Many, many thanks for your time,

    John

    Wednesday, September 25, 2013 6:38 PM

Answers

All replies

  • Hello John,

    What for error messages do you get in SQL Server ErrorLog file and Windows Eventlog?


    Olaf Helper

    [ Blog] [ Xing] [ MVP]

    Wednesday, September 25, 2013 6:43 PM
  • System Eventlog shows Event ID 7036 just after midnight:

    The MSSQLSERVER service entered the stopped state.

    Then a couple seconds later I get another 7036:

    The SQLSERVERAGENT service entered the stopped state.

    I have to connect to the server and start the service once these show up.

    Please see here for the SQL Server ErrorLog file which has this on the bottom:

    2013-09-25 00:11:02.27 spid78    Stack Signature for the dump is 0x6F952303
    2013-09-25 00:11:02.27 spid78    SQL Server Assertion: File: <logscan.cpp>, line=3176
    Failed Assertion = '(m_lastLSN == NullLSN) || (m_lastLSN > m_curLSN)'.
    2013-09-25 00:11:02.28 spid78    Error: 3624, Severity: 20, State: 1.
    2013-09-25 00:11:02.28 spid78    Error: 3314, Severity: 21, State: 4
    2013-09-25 00:11:02.28 spid78    Error while undoing logged operation in database 'tempdb'. Error at log record ID (422:16:247)..
    2013-09-25 00:11:02.28 spid78    Error: 3449, Severity: 21, State: 1
    2013-09-25 00:11:02.28 spid78    An error has occurred that requires SQL Server to shut down so that recovery can be performed on database ID 2..


    • Edited by Kanaloa98 Wednesday, September 25, 2013 9:11 PM
    Wednesday, September 25, 2013 7:17 PM
  • Also seeing these in the Windows Application Log:

    Event ID 17502 - Error: 605, Severity: 21, State: 1
    Attempt to fetch logical page (1:9085) in database 'tempdb' belongs to object '1207634970', not to object '#AnchorText_________________________________________________________________________________________________________000200009097'.

    Event ID 17502 - Error: 7987, Severity: 22, State: 1
    A possible database consistency problem has been detected on database 'Hookahi_Search_DB'.  DBCC CHECKDB and DBCC CHECKCATALOG should be run on database 'Hookahi_Search_DB'.

    Event ID 17505 - 17066 :
    SQL Server Assertion: File: <logscan.cpp>, line=3176
    Failed Assertion = '(m_lastLSN == NullLSN) || (m_lastLSN > m_curLSN)'.

    Event ID 17505 - 18052 :
    Error: 3624, Severity: 20, State: 1.

    Event ID 17502 - Error: 3314, Severity: 21, State: 4
    Error while undoing logged operation in database 'tempdb'. Error at log record ID (422:16:247).

    Event ID 17502 - Error: 3449, Severity: 21, State: 1
    An error has occurred that requires SQL Server to shut down so that recovery can be performed on database ID 2.

    Thursday, September 26, 2013 6:03 PM
  • You can try one of the later hotfix packages here Cumulative list of the hotfixes that are available for SQL Server 2000 SP4.

    But SQL 2000 is out of support, so you might need to migrate those databases to a supported version of SQL Server.

    David


    David http://blogs.msdn.com/b/dbrowne/

    • Proposed as answer by Allen Li - MSFT Thursday, October 3, 2013 8:41 AM
    • Marked as answer by Shanky_621MVP Friday, October 11, 2013 5:22 AM
    Thursday, September 26, 2013 6:18 PM