All,
Has anyone run into any problems with Profiler for Server 2k5 killing their applications?
We have a clustered Production environment (active/passive). The database is connected to via an MS Access ADP file which in turn the end users open via load balanced Citrix Servers (there are 3 or 4 in the farm). Things got a little slow in the past week so I went to run Profiler and PerfMon on the SQL Server.
PerfMon caught some spikes in the Disk Queue Length & Pages per Seconds, but it looked like normal intermintant activity, nothing persistant that could be causing huge problems. Profiler (I filtered it down to a specific database and looked for duration of 30 seconds or longer) wasn't finding any queries which met my criteria.
However, every time I started Profiler --whether or not PerfMon was running, whether I was logging to a trace table, trace file or not logging at all -- All the people logged on through the ADP had problems. The application froze up, screens when blank, queries and reports timed out.
We verified with our system admins that it wasn't a network issue, everything seemed to still work fine through the backend (though we weren't running complex queries), but nothing through the front end worked until Profiler was stopped.
Does anyone know of any issues that might have caused this scenario?
Thanks in advance.
Catadmin - MCDBA, MCSA
"No, no. Yes. No, I tried that. Yes, both ways. No, I don't know. No again. Are there any more questions?"
-- Xena, "Been There, Done That"
Has anyone run into any problems with Profiler for Server 2k5 killing their applications?
We have a clustered Production environment (active/passive). The database is connected to via an MS Access ADP file which in turn the end users open via load balanced Citrix Servers (there are 3 or 4 in the farm). Things got a little slow in the past week so I went to run Profiler and PerfMon on the SQL Server.
PerfMon caught some spikes in the Disk Queue Length & Pages per Seconds, but it looked like normal intermintant activity, nothing persistant that could be causing huge problems. Profiler (I filtered it down to a specific database and looked for duration of 30 seconds or longer) wasn't finding any queries which met my criteria.
However, every time I started Profiler --whether or not PerfMon was running, whether I was logging to a trace table, trace file or not logging at all -- All the people logged on through the ADP had problems. The application froze up, screens when blank, queries and reports timed out.
We verified with our system admins that it wasn't a network issue, everything seemed to still work fine through the backend (though we weren't running complex queries), but nothing through the front end worked until Profiler was stopped.
Does anyone know of any issues that might have caused this scenario?
Thanks in advance.
Catadmin - MCDBA, MCSA
"No, no. Yes. No, I tried that. Yes, both ways. No, I don't know. No again. Are there any more questions?"
-- Xena, "Been There, Done That"