Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
Westley
Posting Yak Master
229 Posts |
 Posted - 2005-12-28 : 19:41:23
|
| Hi all, Just wondering for a normal production server, will you guys tend to setup SQL alerts? like using some default SQL alerts with some user defined ones? (eg, log level over some percentage for some DBs etc etc) and some Severity alerts (like from 19 to 25), would just like to know if you have all these in place, will that put more overhead to the server as it needs to keep track of them, so the more you have, the slower the server will be? but without those alerts, what will be the best way to monitor those events (for more then 30 servers of cos).Thanks, |
|
|
sachinsamuel
Constraint Violating Yak Guru
383 Posts |
Posted - 2005-12-28 : 23:48:51
|
| We have alerts and operators in all our production server. While configuring them we specify the criteria for them to get execute. Like production servers disk or raid getting out of space, Any Fatal error. They help you to monitor the production server even if you are not in your lab. In case u don't have alerts on place, you can create more steps in your jobs and in the advance tab, call them when it fails.RegardsSachinDon't sit back because of failure. It will come back to check if you still available. -- Binu |
 |
|
|
Westley
Posting Yak Master
229 Posts |
Posted - 2006-01-03 : 01:22:30
|
| Thanks Sachin, Guess my question is really will that have a big performance hit if you do that, as if u have like 30 alerts, some on log space size, and some on major sql issues, its understandable to have alerts on major sql issues, but isn't that the more you have, the slow the server will be? as its just keep adding the overhead to the server?it just like traces, the more you have, the more resources it will use therefore slow the performance will be.....?? |
|
|
|
sachinsamuel
Constraint Violating Yak Guru
383 Posts |
Posted - 2006-01-03 : 01:39:50
|
| There is not big performance hit as such. But alerts are set on events which need your attention like disk running out of space. Or Backup failed. But if you want you can also go for common events.RegardsSachinDon't sit back because of failure. It will come back to check if you still available. -- Binu |
|
|
|
Westley
Posting Yak Master
229 Posts |
Posted - 2006-01-03 : 04:13:29
|
| Thanks for that, I always though it will be a big performance hit as the trace and the alert will put more overhead to the server. But its good to hear that its not, will try out some of them and see what happens. |
|
|
|
|
|
|
|
|
Alerts monitored