December 14, 2010 at 10:47 am
I have Quest Spotlight monitoring several servers, which runs a server-side trace on each instance monitored. Every time a SQL server gets rebooted, Spotlight creates and starts a new trace without removing the old trace. After several months, I have between 4 and 8 server-side traces running that were all created by Spotlight for every instance monitored on all servers.
I opened a ticket with Quest, but it's going nowhere. Quest claims they only run one trace per instance and that it's a Microsoft issue if I have old traces still running after a reboot. Here's the exact quote from Quest, "That is a sql server issue, you might want to contact Microsoft on that one since the trace is a sql server trace run by the [Spotlight] user. Normally when the server reboots all traces should go with it, spotlight creates another one [because] the old traces are inactive, made so by the reboot. "
Is this true or should Quest be cleaning up after itself?
December 14, 2010 at 11:14 am
Restarting the SQL service, whether from just restarting the service or from rebooting the server/VM, dumps all traces except the default trace (usually trace ID 1).
I'd take a look in SQL Agent and see if there are any jobs set to run At Startup, that are starting traces, or any stored procedures in the master database that have been configured to run on startup. If it's not in either of those places, then it's outside of SQL Server altogether, possibly in something in the Windows startup folder.
- Gus "GSquared", RSVP, OODA, MAP, NMVP, FAQ, SAT, SQL, DNA, RNA, UOI, IOU, AM, PM, AD, BC, BCE, USA, UN, CF, ROFL, LOL, ETC
Property of The Thread
"Nobody knows the age of the human race, but everyone agrees it's old enough to know better." - Anon
December 15, 2010 at 7:59 am
We had windows updates go in last night so all servers rebooted and flushed all traces as you said they would. Spotlight then created one new trace for all monitored servers. Finally, the Spotlight diagnostic server rebooted. When it came up, it created a new trace for all monitored servers again, so now each has two traces besides the default. Had the order of reboots been reversed, I'd have one new trace per server.
Thanks for you help and explanation. I'll continue to monitor and diagnose why the number of traces increased to the level they did in the past. I have a hunch it has to do with the diagnostic server rebooting or service restarting. 😉
Viewing 3 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply