[mrtg] Unknown SNMP var hrStorageAllocationUnits

[Stefan Midjich]

I get these errors for all the hrStorage OIDs. Followed by this.

2012-04-05 10:00:45: WARNING: Expected a number but got '1 day, 18:14:44'

I've seen this before on the list but not found a solution for my own
situation.

I have LoadMIBs: /usr/share/mibs/netsnmp/UCD-SNMP-MIB,
/var/lib/mibs/ietf/HOST-RESOURCES-MIB in my mrtg.cfg.

I first set this up at home for a Debian 6 system and it worked fine, then
I simply copied the mrtg.cfg to work and changed what needed to but kept
most of the OIDs intact. Only in some cases did I have to change the end of
the OID but mostly snmpwalk showed they were identical since I was
monitoring Debian 6 at work too.

However, at work it all broke down as if I had taken a fish out of water.
These errors never showed up at home.

So nothing is working now and I'm getting a ton of these errors.

$(snmpwalk -Os -c mycommunity -v 1 myhost hrStorage) shows all the OIDs
intact and working so I'm not sure why mrtg is giving so much grief.

I was hoping this was a known issue. At home I run 2.17.3 and have no
problems, at work I run 2.16.3 and have all the problems.

Besides the errors with hrStorage I can't get the CPU graphs nor the
traffic graphs to even update. I'm pasting the configuration here for one
host, this repeats itself 8 times with different hostnames but since all 8
machines are identical Debian 6 installs they all seem to have the same
OIDs so far.

http://pastebin.com/xEGxRang

Every single OID has been checked using snmpwalk and it shows up just fine
but in mrtg there are issues.

-- 
Hälsningar / Greetings

Stefan Midjich
[De omnibus dubitandum]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.oetiker.ch/pipermail/mrtg/attachments/20120405/604d058a/attachment.htm