SQL Agent stops unexpectedly (SQL 2000 SP1)

  • Our SQL 2000 machine had SP1 installed recently and since then the SQL Agent stops with no warning and with no apparent pattern except mostly on the hour or half hour. e.g. 7:00 AM, 12:30 PM, 9:30 AM etc

    None of the tasks run by the agent have been executed at all the times that the Agent has stopped.

    The windows event viewer has an error logged like:

    The SQLServerAgent service terminated unexpectedly. It has done this 1 time(s). The following corrective action will be taken in 180000 milliseconds: Restart the service.

    Our server has 4GB of RAM and the /PAE switch and AWE enabled in SQL configuration. THe machine is using Win2K AS and SQL 2000 EE.

    Note: The problem is not the same as the problem with multiple emails in an operator causing SQL agent to stop.

    Has anyone else experienced something like this? Without giving anyone pre-conceived ideas, I would like to know if anyone can describe something similar and what symptoms you may have seen.

  • Im running the same software config on two boxes, one an 8 processor box with 8G of ram, the other 4 processors 4G, have not seen this problem.

    Andy

  • Neil or anyone,

    Can you tell me more about "Note: The problem is not the same as the problem with multiple emails in an operator causing SQL agent to stop"?

    I have not heard of this and searches in TechNet yield nothing so, I am curious. Thanks in advance.

    David

    David

    @SQLTentmaker

    “He is no fool who gives what he cannot keep to gain that which he cannot lose” - Jim Elliot

  • Email bug is described in KB article Q302892.

  • Andy Warren, Could you tell me if you have heavy processing (e.g. greater than 80% across all CPU's) and a SQL agent task start off during this condition? Thanks.

  • Neil, Im not having this problem, was just commenting. I try to avoid 80% anyway!

    Andy

  • Are you running replication agents (epecially merge and log reader agents)?

    Alternatively, do you have any DTS packages or Command Executable Jobs which are scheduled during these times?

    Also, are you using SQLMail?

    All of these run in process with SQLAgent and could be causing a resource allocation in SQLAgent. Since you are running AWE what is your max server memory set to?

  • Thanks for the thoughts.

    No replication is in use.

    The jobs scheduled to run do not appear to be the cause as no one job has been running at all times when the agent has stopped. Likewise for DTS, xp_sendmail and xp_cmdshell.

    There is a known bug to do with operators using multiple email accounts, but I have been able to rule this out with the work around.

    The server has 4 GB of RAM and 3776 MB is reserved for SQL Server (leaving about 256 MB for OS). The only other application we run is Arcserve backup software. This steals pages from SQL if memory is dynamic.

    I have also ruled out Arcserve since it has never run when the problem occurs.

    I have noticed that the problem occurs during times of heavy use.

    Let me know if you see something similar.

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

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