On 10/18/06, joel at carnat.net <joel at carnat.net> wrote:
Hi,
I wrote a script to get alert about CPU loads on a windows server using SNMP (no I don't want to / can't install the hobbit client on the server ;). The status is returned OK (green before 50%, orange from 50 to 75 and red over 75).
The thing is, there is no graph done (yet). I would like a graph that would show the load average variation.
(how) can the script be used to fill a RRD database ? Should the load average value be returned in a special variable ?
I'm doing something similar - SNMP queries against a Compaq system to produce a report like this:
CPU 1 min 5 min 30 min 60min &green 0 14% 9% 9% 9% &green 1 11% 7% 9% 9%
I changed the [rrdstatus] entry in server/etc/hobbitlaunch.cfg to include the --extra-script & --extra-tests entries as described in the hobbit_rrd man page.
If it gets through the mail system, my collection script is attached. It's not rocket science, but it seems to get the job done.
That pokes the data into an RRD, then it's just a question of setting up the graph config to show what you want to see.
Ralph Mitchell