Question about SQLAgent - Alert Engine

[DeadManWalking]

Just today we were seeing out SQL Server(2k) staying between 95% and 100% CPU utilization. So I ran the following query:

select hostname, program_name, loginame, cpu, physical_io, memusage,
status
from master.<name>.sysprocesses
order by cpu desc

and the SQLAgent - Alert Engine CPU was running at 611100, physical io at 28 and memory usage at 0. Is this normal for the Alert Engine to be the hardest on the CPU?

[MAK]

It is normal.

BTW. when did you recycle your box recently?

[DeadManWalking]

About 10 days ago.

We're also looking into redoing the indexes on a more regular basis. We do alot of inserts and updates(At least 10,000 to 20,000 inserts a day). What time frame would you say we should re index the tables?

[MAK]

Do the reindexing during off hours. first try it on a sat and sun and see how long does it takes, then schedule it.

[DeadManWalking]

Sorry, I should have worded it better. Do you think we should do it weekly? Daily? I've really havn't read any good articles or docs. giving a good idea on frequency of the reindexes...

[MAK]

I would recommend Daily for such amount of transactions.

[DeadManWalking]

Sounds good. Thank you for the great advice.

[MAK]

remember. run update stats or do using maintenance plan.