I have 350 machines that i monitor.
Every linux box has 6 partitions. Is it possible to build up a rrd that will contain the 6 partitions per machine?
I want a nice graph that gives me hda1, hda3 hda4 hda5…
Mon Jun 13 12:05:24 2005
**INCOMING PERFDATA:
LASTCHECK=1118657115
HOSTNAME=befrs999c1
SERVICEDESCR="DISK1"
SERVICESTATE="OK"
OUTPUT=“DISK OK - free space: /boot 120 MB (88%):“
PERFDATA=” /boot=17MB;116;126;0;136”
Yes, you can put several items in one rrd file. Here is a snip from my nagiostat.conf file that shows the graph was created to store “rta” round trip average for a ping, plus “pktloss” packet loss.
RRDCreateTemplate ping_5min --step 300 DS:rta:GAUGE:600:0:5000 DS:Pktloss:GAUGE:600:0:100 RRA:AVERAGE:0.5:1:396 RRA:AVERAGE:0.5:6:336 RRA:AVERAGE:0.5:24:480 RRA:AVERAGE:0.5:234:480
When the nagios check_fping command is ran, it gives the values for both rta and pktloss. The output of the command is sent to nagiostat, it then parses the output, and stuffs the data in the rrd. So if your check_command that looks at the partitions can give you results for all 6 partitions, then create the line in the nagiostat.conf that will accomplish the create/parsing.
hmmz is there somewhere a good example with extra comments …
i cant get it to work
the man pages for rrdtool should help quite a lot.
Luca
Ok I’ve got the check_memory working 
i still have problems with the hda graphs
the output is service check DISK1
"DISK OK - free space: /boot 120 MB (88%):"
the output of serice check DISK2
OUTPUT=“DISK OK - free space: / 2661 MB (89%):”
RRDCreateTemplate hd_5min --step 300 DS:hd1:GAUGE:600:0:5000 DS:hd2:GAUGE:600:0:5000 RRA:AVERAGE:0.5:1:396 RRA:AVERAGE:0.5:6:336 RRA:AVERAGE:0.5:24:480 RRA:AVERAGE:0.5:234:480
But now how do i get nagiostat to understand that i only want 88% for hda1 and 89% for hda2
if you are usig nagiostat there is a description in the nagiostt.conf file about how to pass the paramters to the rrds and how to extract them from the perfdata.
Luca