Starting one named SQL Server service stops/crashes another

  • Hi all

    We currently have a machine running a named instance of SQL Server 2008 and need to create a new named instance on the same machine (the separation is needed due to client data from different countries). In the past this has been no problem and we have multiple instances on other machines running with no issue.

    In this case however after having installed the new instance when starting up the new services, it either stopped (or crashed) the existing instance service "SQL Server (Original)". Noticing this I restarted the service for "SQL Server (Original)" service, but then realized that it then stopped the new one "SQL Server (New)".

    Obviously there is some conflict somewhere, but we have never had this issue previously. I thought it may have been a TCP issue but having looked at the Network Configuration both seem to be running with Dynamic Ports with no clash in the port number.

    I've not had much luck with Google but may be searching for the wrong terms - has anyone out there seen a similar problem and/or come up with a solution?

    Many thanks for any help.

  • What does the error log of the system being shut down say? Are there any entries in the Windows logs suggesting possible problems?

    "The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
    - Theodore Roosevelt

    Author of:
    SQL Server Execution Plans
    SQL Server Query Performance Tuning

  • Grant, thanks for your reply.

    I neglected to mention in the earlier post that the SQL Server log showed no error at all. There was simply the last valid entry of one log (a log backup being taken) then the next log started with the usual instance startup entries. I did however receive an agent alert email notification for a Severity 25 fatal error: The <InstanceName> service terminated unexpectedly.

    For the windows logs:-

    The security log showed the services for the new named instance being granted their various privileges, but no untoward errors at all.

    The system log showed the various services of the new instance "entering the running state", 2 seconds later the next message showed the SQL Server Agent (Original) service entering the stopped state, following 1 sec later "The SQL Server (Original) service terminated unexpectedly. It has done this 2 time(s)." - this was the last message before I restarted the (Original) service. Once this service started it was followed immediately by "The SQL Server (New) service terminated unexpectedly. It has done this 1 time(s)."

    The application log showed the installer going through its thing and then the new named instance starting up. Here the last service to start up was the Reporting Services followed by "The MSSQL$<Original> service terminated unexpectedly."

    Essentially all I can see is that the new instance starts up and then the old one stops (and vice versa). Unfortunately there does not seem to be a lot to go on...

  • Were they installed incorrectly, sharing the same systems dbs? I've never tested the impact of this so I cant confirm what impact this would have.

  • I'm not sure. You may need to open a case with Microsoft. I'd say it sounds like you've set them up on the same ports, but you said you checked that. If it's not that and there's no indications in the error logs getting in touch with Microsoft would be my next step.

    You could try running pssdiag. They're probably going to have you do that anyway. That might capture something you can spot.

    "The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
    - Theodore Roosevelt

    Author of:
    SQL Server Execution Plans
    SQL Server Query Performance Tuning

  • You could consider running File Monitor (http://technet.microsoft.com/en-gb/sysinternals/bb896642.aspx) and Process Monitor (http://technet.microsoft.com/en-gb/sysinternals/bb896645.aspx) to see if the new instance is trying to access any files that belong to the original instance.

    Regards,

    Phil

  • Failing any of the previous suggestions, I'd start looking at available resources... Does the server have enough RAM to run all the existing instances? Is there perhaps a disk space issue somewhere ? Admittedly, this is more of a "grasping at straws" thing, but outside of going to MS, there might not be any easy ways to trap the problem.

    Steve (aka sgmunson) 🙂 🙂 🙂
    Rent Servers for Income (picks and shovels strategy)

  • The most obvious cause is that both instances are sharing a resource that should not be shared. The advice about using File Monitor and Process Monitor to help detect this is worth looking at.

    You say you are using SQL 2008 (not SQL 2008 R2). This version uses 8.3 style filename references for some of its folders, and I remember a situation a few years back where the installer plugged in the wrong 8.3 reference for a folder. This resulted in two instances using the same folder when looking at 8.3 format filenames, while the long format filenames were correct and distinct. Needless to say only one instance at a time would work. I could not reproduce the problem so did not report it.

    Original author: https://github.com/SQL-FineBuild/Common/wiki/ 1-click install and best practice configuration of SQL Server 2019, 2017 2016, 2014, 2012, 2008 R2, 2008 and 2005.

    When I give food to the poor they call me a saint. When I ask why they are poor they call me a communist - Archbishop Hélder Câmara

Viewing 8 posts - 1 through 7 (of 7 total)

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