0

I just upgraded a simple Ubuntu webserver from 9.10 to 10.04.1 and our nagios claims UNKNOWN for the single snmp check we run against it. It was working previously, so this baffles me.

I installed the plugins on my desktop and get this output when I run it (obfuscated where appropriate:

$: ./check_snmp_storage.pl -H $HOSTNAME$ -C $pword$ -m / -w 90 -c 95
ERROR: Description/Type table : No response from remote host $HOSTNAME$`

So I installed nagios plugins on the webserver and get the same output, unless I use localhost for $HOSTNAME$.

The snmpd.conf is pretty trivial, and I wonder if that's the problem?

syslocation  "ESX Server"
syscontact  admin@example.com
rocommunity  pword
disk  /
load  30 30 30
jldugger
  • 14,342
  • 20
  • 77
  • 129

1 Answers1

3

Check the file /etc/default/snmpd. By default on debian/ubuntu this file has an option which restricts snmp to the local host. Here is a diff from the stock config. I have seen this get reverted back to the stock config during an upgrade.

--- a/default/snmpd
+++ b/default/snmpd
 # snmpd options (use syslog, close stdin/out/err).
-SNMPDOPTS='-Lsd -Lf /dev/null -u snmp -I -smux -p /var/run/snmpd.pid 127.0.0.1'
+SNMPDOPTS='-Lsd -Lf /dev/null -u snmp -I -smux -p /var/run/snmpd.pid'

If the above isn't the issue, double check the firewall on the host if it has one.

Zoredache
  • 130,897
  • 41
  • 276
  • 420
  • That may well be it. I don't recall changing it, so it wouldn't have triggered a diff on upgrade. I'll check when I get in tomorrow. – jldugger Aug 13 '10 at 00:51
  • That was it. Reviewing the changelog shows Debian patched in that in 2006. If I get time I might check etc-keeper and verify that it wasn't there before. – jldugger Aug 13 '10 at 18:27