What is the best way to configure performance alert

[Printaporn_1]

In some event for ex. CPU reach 90% then send alert via mail.

How to implement this ? using measure , glance , gpm , advisor ?

thanks in advance,

[Thierry Poels_1]

hi,

# Default alarmdef definition file is /opt/perf/newconfig/alarmdef.
# System-specific alarmdef file used by alarmgen daemon is in /var/opt/perf.
# Additional examples exist under /opt/perf/examples/mwaconfig.

regards,
Thierry

[Sajith V Mannadiar]

Hi,

Install and configure BigBrother monitoring tool.

That can send you alerts for CPU, disk , network outages.. It's too good.

Download that from http://bb4.com

Need any further help on this, mail me

Regards,
Sajith

[Fred Martin_1]

Here's a script that is probably more complex than you need, but I use it. It runs in root cron every 1/2 hour.

If a file system is found to be over some limit, it sends an email and a page (via email). The email is only sent once; the page would be sent every 1/2 hour until the situation is resolved.

Also not in the script, certain file systems are ignored - because I expect them to be over the limit. They are non-production file systems anyway.

Script should be attached.

[Fred Martin_1]

I just noticed you were looking for CPU usage, not disk space. The script could be modified to suit.