October 22, 2010 at 1:46 pm
I was trying to gather traces to do a database tuning.
Using "Tuning" template in SQL Profiler I started a trace to gather events for 1 hour. Data gathered was stored in a table of a Debug database.
However, after several minutes the server went to it's knees and users started complaining.
I stopped the trace and the server went back to normal.
I understand that running Tuning Advisor is resource-demanding and should only be run off peak hours, but should traces also be avoided?
Please help!
October 23, 2010 at 2:17 am
It is not recommended to run Profiler trace on a Production instance. Instead run a server side trace using as appropriate. An example here[/url]
Pradeep Adiga
Blog: sqldbadiaries.com
Twitter: @pradeepadiga
October 23, 2010 at 5:22 am
You should absolutely not run profiler on a production system as was mentioned. In addition, you should output your trace to file, not a database and especially not a database on the same server. You can use fn_trace_gettable to read the files and input them into a table.
Here is an excellent book on Profiler. Brad's the man.
"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
October 23, 2010 at 9:46 am
Thank you Pradeep and Grant for the replies and tips.
I was not aware that Profiler stages the trace on C: drive. In addition, to ad injury to insult, I found that a daily C: drive defrag was scheduled for that same time on that server! I will create a server side trace instead. And of course, have the defrag rescheduled.
Thanks!
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply