Please start any new threads on our new site at https://forums.sqlteam.com. We've got lots of great SQL Server experts to answer whatever question you can come up with.

 All Forums
 SQL Server 2000 Forums
 SQL Server Administration (2000)
 Urgent Problem - SQL Agent

Author  Topic 

elancaster
A very urgent SQL Yakette

1208 Posts

Posted - 2004-12-21 : 04:55:03
Why does everything go wrong as soon as we go on holiday for a few days?

We are running SQL2000 on Win2K, 4 processors (4gb), as part of a legato cluster.

A few days ago they found that the cpu was running at a constant 100% which obviously caused major problems to the performance. They stopped SQL Agent and the cpu fell to normal again. Whenever they tried to start it again it max's out. It seems as though something is running but we cannot tell what. When Agent is on there is nothing showing as running.

They rebooted the server to try and kill whatever it was trying to run. This of course meant that it failed over to the secondary server. When they started Agent again it max'd out again, as though still trying to carry on whatever it is doing.

I wasn't here at the time, no-one ran a trace at the time to find what it is running but I cannot start Agent again now until we are very confident that it will not kill everything (not my call). Of course Agent being off, means that the back-ups are not running and the hard-drive is filling up fast (log file). This is where the urgency comes in. I'm going to run a manual back-up tonight if we can't sort it out before then.

Does anyone please have any suggestions of what might be causing it / how I can sort it? I was assuming it is something trying to run in the background (a hanging dbcc job or something maybe?), but when it didn't clear it by rebooting I'm not sure. Because of the clustering could the secondary server have just carried on where the primary left off? If it something running, is there a way of clearing out hanging jobs without rebooting?

help!

thanks

Kristen
Test

22859 Posts

Posted - 2004-12-21 : 12:40:52
Make a note of which jobs are enabled (assuming SOME are disabled)

Mark all jobs as disabled

Start SQL Agent

Enable some jobs - e.g. Backups. let them run to completion.

Enable some more jobs, one by one, and see what breaks the server!

Kristen
Go to Top of Page

jen
Master Smack Fu Yak Hacker

4110 Posts

Posted - 2004-12-21 : 22:01:17
have you tried looking at the current activity monitor?

how's the other node performing? does it encounter the same issue with the "problem" node?

--------------------
keeping it simple...
Go to Top of Page

MichaelP
Jedi Yak

2489 Posts

Posted - 2004-12-21 : 22:19:34
The above suggestions are great. I'd probably fire up Profiler and then enabled jobs one-by-one to see what's going on.

Michael

<Yoda>Use the Search page you must. Find the answer you will.</Yoda>
Go to Top of Page
   

- Advertisement -