SQL Agent Auto-restart in a cluster

  • How do I get the SQL Agent service on a cluster to autorestart?

    When the server fails over, the SQL Service restarts, but the agent does not. When I go the Service Properties in EM the option is grayed out. When I go to the Services MMC both MSSQL and the agent service show as manual. MSSQL is started, but the agent is not.

    Also, in cluster administrator the Agent is in the same group as the rest of the SQL stuff and lists as dependant on the SQL service, and the restart option button is selected.

    Any help on this topic would be appreciated.

  • SQL Server Agent maybe fail to start after fail-over. Check SQL Server Agent log (Display error log) from EM and review machine application/system logs too.

  • Check your sqlagent-settings in the cluster-administrator. On the "advanced"-tab you'll find the options.

    Johan

    Learn to play, play to learn !

    Dont drive faster than your guardian angel can fly ...
    but keeping both feet on the ground wont get you anywhere :w00t:

    - How to post Performance Problems
    - How to post data/code to get the best help[/url]

    - How to prevent a sore throat after hours of presenting ppt

    press F1 for solution, press shift+F1 for urgent solution 😀

    Need a bit of Powershell? How about this

    Who am I ? Sometimes this is me but most of the time this is me

  • Thanks for the input.

    However, I get no errors in any logs. It's not that the service won't start, it is that it does not auotmatically restart at failover. I can manually start it no problem.

    SQL Agent settings in Cluster admininistrator's advanced tab is already set to Restart.

    Anything else that you can suggest would be appreciated

  • In EM, check the Properties of the SQL Server in question. Go to the Processor tab, and check to see if the "Use Windows NT fibers" check box is checked. You may have to uncheck it and fail over. I had this same issue, and it went away after I unchecked this box, which I think is defaulted to being checked when Enterprise Edition is installed. It turns out that it is a timing issue, like so many of the problems with fibers and multi-threaded applications.

    Hope this works for you!

  • Thanks for the reply.

    I looked at the EM's Server properties, and "Use NT Fibers" is already cleared.

  • It sounds to me that this option may not be Cluster aware (either by bug or by design)and must be changed via the SQL Server Service Manager while the SQL Server Cluster resource is OFFLINE. The old "bludgeon method" I would call it.

    Assuming this is an Active/Passive cluster, I would do the following:

    1. On the Passive Node, take SQL Server offline. Set the Autostart option via the SQL Server Service Manager (on the taskbar).

    2. Bring the SQL Server Cluster resources back online

    Now Failover to the Passive Node you just administrated and check the result. Repeat, lather and rinse for the other node as well.

    I'd be interested in learning if this requires just the SQL Server Agent Cluster Resource to be taken OFFLINE or if all resources must be taken OFFLINE.

    Stuart

    "Chance is always powerful. Let your hook be always cast. In the pool where you least expect it, will be a fish" - Ovid

  • quote:


    Assuming this is an Active/Passive cluster, I would do the following:

    1. On the Passive Node, take SQL Server offline. Set the Autostart option via the SQL Server Service Manager (on the taskbar).

    2. Bring the SQL Server Cluster resources back online

    Now Failover to the Passive Node you just administrated and check the result. Repeat, lather and rinse for the other node as well.

    I'd be interested in learning if this requires just the SQL Server Agent Cluster Resource to be taken OFFLINE or if all resources must be taken OFFLINE.


    Good thought, I do not yet have the second node of the cluster running. When I do I will try as you suggest. I'll let you know the results.

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

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