Issue with check_snmp_storage.pl and Solaris automount

[dsolberg]

Greetings,

I currently an running volume checks against our Solaris 11 servers using the check_snmp_storage.pl. We currently do not have that many volumes on the servers, less than 10, we also use a number of automounts for mostly home directories. The issue I am running across is that the automounts will not unmount due to the fact that with the current configuration for the automount will only dismount after 10 minutes of inactivity and the snmp process "pings" the automount directory from the snmp check, blocking the dismount.

So the question is, and looking for guidance, is there a way of grouping checks? So if I could extend the time in between SNMP checks so that there was a period of inactivity on the automount, it would be able to dismount naturally. Of course, there is also the option of checking the disk by ssh, which I have had issues in the past as well. With Solaris, there is a hard set limit of a maximum 10 connections by ssh before it starts to refuse connections.

Any insight would be greatly appreciated.

Thanks, Don

[jdalrymple]

Don
It looks like the check_snmp_storage.pl plugin will accept some custom arguments to ignore certain mountpoints:

Code: Select all

-m, --name=NAME
   Name in description OID (can be mounpoints '/home' or 'Swap Space'...)
   This is treated as a regexp : -m /var will match /var , /var/log, /opt/var ...
   Test it before, because there are known bugs (ex : trailling /)
   No trailing slash for mountpoints !


  the -m option allows regexp in perl format :
  Test drive C,F,G,H,I on Windows       : -m ^[CFGHI]:
  Test all mounts containing /var       : -m /var
  Test all mounts under /var            : -m ^/var
  Test only /var                        : -m /var -r
  Test all swap spaces                  : -m ^Swap
  Test all but swap spaces              : -m ^Swap -e

Did you fiddle with that at all?

[dsolberg]

Thank you for the reply.

I did look at that, I am performing the tests all with the -m parameter, the automount is "/admin", in this case and I am not doing any check on the root volume, just volumes such as "/u30", "/u31", etc...

Tracing shows that all mountpoints are being returned regardless of the paramters and the check_snmp_storage.pl is parsing out the information in the parameter as requested.

Thanks!

[jdalrymple]

So your command looks something like

Code: Select all

check_snmp_storage.pl -H $HOSTADDRESS$ -C $_HOSTCOMMUNITY$ -m ^/admin -e -w 70% -c 80%

or some such?

[dsolberg]

More in the line of:

./check_snmp_storage.pl -H ta01 -C XXXXXXX -w 85 -c 90 -m ^/u31 -f
/u31: 0%used(0MB/59976MB) (<85%) : OK | '/u31'=0MB;50980;53978;0;59976

When I attempt to run the command listed, I receive the following;

./check_snmp_storage.pl -H ta01 -C XXXXXX -m ^/admin -e -w 70 -c 80
ERROR: Size table :Message size exceeded maxMsgSize.

Not sure how I could run with one exclusion and still be able to test other mountpoints, does not look like I can run multiple parameters on the check command, one to include, and one to exclude. I could be confused, that happens too!

[jdalrymple]

-e just inverts the -m pattern as I understand it. It could go either way - either group the ones that do match, or the ones that don't match. It does work as expected for me:

Code: Select all

[jdalrymple@somehost libexec]$ ./check_snmp_storage.pl -C rocomm -H 192.168.0.1 -m / -w 70 -c 80
/sbin: 35%used(16MB/45MB) /tmp: 0%used(0MB/15MB) /root: 0%used(0MB/48MB) /var: 13%used(8MB/62MB) /etc: 10%used(5MB/48MB) /flash: 8%used(619MB/7391MB) /bin: 36%used(6MB/16MB) /usr: 60%used(528MB/875MB) /: 88%used(3MB/3MB) (>80%) : CRITICAL
[jdalrymple@somehost libexec]$ ./check_snmp_storage.pl -C rocomm -H 192.168.0.1 -m /var -e -w 70 -c 80
/sbin: 35%used(16MB/45MB) Swap space: 0%used(0MB/0MB) /tmp: 0%used(0MB/15MB) Real memory: 31%used(316MB/1008MB) /root: 0%used(0MB/48MB) Physical memory: 31%used(316MB/1008MB) /etc: 10%used(5MB/48MB) /flash: 8%used(619MB/7391MB) /bin: 36%used(6MB/16MB) /usr: 60%used(528MB/875MB) /: 88%used(3MB/3MB) (>80%) : CRITICAL