Cpu Usage on Monitored Server

[goldmedal]

Hi

We have noticed a serious problem with the monitoring on one of our production servers, the cause of which is SQL Response.

Our understanding of the product that the impact on the monitored servers is minimal, however through its current interaction it is using between 10-15% of the monitored servers CPU at regular intervals to execute the following code.

DBCC TRACEON WITH NO_INFOMSGS
DECLARE @dbccstat TABLE
(
TraceFlag INT,
[Status] INT,
[Global] INT,
[Session] INT
)
INSERT INTO @dbccstat EXEC('DBCC TRACESTATUS(1204, -1) WITH NO_INFOMSGS')
DECLARE @status INT
SELECT @status = Status FROM @dbccstat
IF @status = 0
BEGIN
DBCC TRACEON( 1204, -1 ) WITH NO_INFOMSGS;
END
EXEC sp_readerrorlog

Can you advise on this? As in the current setup this renders sql response un-usable for us in a production environment as CPU usage is at a premium.

Thanks

[kevriley]

Could the size of your error logs be an issue?

Are you recycling them?

Do they take a long time to open in SSMS?

[goldmedal]

Hi

Thanks, i've had a look at the recycling schedule of the logs and this wasn't set up so ive configured it now.

Ive also reviewed the undocumented proc the code is executing to read the error log just to get a better handle on what it was actually doing as i wasn't familiar with the proc.

Hopefully this should resolve the issue of increased CPU load.

Thanks again for your help!