Hit fatal error server is terminating - what do I do next

  • dogramone

    Hall of Fame

    Points: 3215

    Hi all, recently a newly setup (12 months!) SQL Server 2017 instance shut itself downs due to the subject line message.

    I bought the server back up and ran DBCC checkdb in check only mode and no errors were returned. The system has been up and running for a month now since the crash.

    Firstly should I be concerned?

    Secondly should I be sending the dump log anywhere in MS to get any help or advice?

  • Jo Pattyn

    SSC-Dedicated

    Points: 31258

    You should surely be concerned.

    Is the instance up to date (CU 14?)

  • dogramone

    Hall of Fame

    Points: 3215

    The instance has had no patching so only the installed version without any patches.

    And yes I am concerned.

  • Jo Pattyn

    SSC-Dedicated

    Points: 31258

    As it had no patching, you should patch it to the latest cumulative pack.

  • dogramone

    Hall of Fame

    Points: 3215

    I will definitely be planing upgrade. That will be fun seeing as it is in a vendor app and we have no test bed for changes. So fun dealing with vendors and checking if the upgrades will create any issues.

     

    Is there any other advice or is the patching the most important thing?

  • Theo Ekelmans

    SSCarpal Tunnel

    Points: 4479

    Well,

    I's take a good look at the windows and SQL logs to see what happens just before the crash, and see if you can find a precursor to that error ?

    Like, ram complaining about ECC errors and such?

  • Jo Pattyn

    SSC-Dedicated

    Points: 31258

    If you have a support contract you can send the dump to MS. Perhaps you hit an unknow bug.

    Try to get a testbest and verify your backup/restores.

     

  • dogramone

    Hall of Fame

    Points: 3215

    The good/bad news is all that I can find in any event logs is "The SQL Server (MSSQLSERVER) service terminated unexpectedly. It has done this 1 time(s)." There is nothing else. The dump indicates no CPU or memory pressure. There is nothing written to any of the logs within an hour of this occurring and in the SQL logs there is the backup that ran 6 hours before the crash and nothing more.

    And the great news is the DB backed up successfully and restored successfully onto another server so with DBCC checkdb returning no errors and the data appearing to have no corruption.

    The client has no extended support contract, just what is provided with the licence. So is there any where that I can raise this one directly with MS?

  • SGT_squeequal

    SSCertifiable

    Points: 7046

    if this SQL server is part of a vendor application, i would speak with the vendor before applying any patch. Just to ensure you are covered with your support contract should issues arise 🙂

    ***The first step is always the hardest *******

Viewing 9 posts - 1 through 9 (of 9 total)

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