Cheshire Cat Computing

Is it possible to use the vSphere Management Assistance (vMA) Virtual Appliance, instead of installing the Perl SDK directly on the Nagios host?

You need to have the Perl SDK on whichever machine is running the check_vmware plugin, since it uses the routines in the SDK library.

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

2 (hopefully) quick questions -

1. Can we pass variables to the script for Error/Warning conditions? I'm monitoring an ESX server we're currently using for testing and it's Critical for memory @ 92%, but any value up to and including 99% is absolutely acceptable, as it's for testing.


2. I keep getting emails for a memory warning condition for one particular ESX server (via Nullmailer), but whenever I check the status in Nagios it is fine (and the graph doesn't show any spikes). Could this be a bug in the script? I've checked the mail logs, and the software (Nagios/check_vmware) is initiating the mail [ie. it's not a mail loop/error]

The --crit and --warn parameters allow you to specify thresholds (use --help for all the available parameters and information).

The script doesnt send emails itself; if Nagios is sending a notification, it would be because the script is returning a wanring status. It might be a brief spike in usage that is not long enough to appear on graphs? You might want to increase the max_check_attempts or threshold... check the nagios alert log for the history.

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

Hmm...strange. There is nothing in the nagios.log file for the entire day indicating a WARN or CRIT status was reached for that service, but I've received 2 emails for it today, and 5 emails yesterday.

There's nothing in /var/spool/nullmailer/queue/

The emails can be seen in /var/log/mail.log


Damnit.

Hmm...random mail message stopped by itself a few days ago.

Any reason why check_vmware reports a LUN as having 622GB capacity, while the VI client reports 594? Is it the old kilobyte = 1000 VS 1024 story?

SOunds like it may be the 1024/1000 problem. The check_vmware only reports the number as returned by the API.

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

Its a very nice plugin you have made, and all works very well
I seem to have a little problem that you maybe can help me with.

I cant seem to get the network stats report to work ?
Im using cmd:
./check_vmware --mode=nagios --config=/usr/local/nagios/etc/vc.cfg --guest=Web1 --report=net

It returns nothing :/ all other reports works just fine.

There seem to be some problems with network monitoring (see other threads) which might come down to the level for stats collected on your virtualcentre. Try raising the stats to level 3.

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11

Verion 1.13 posted (see the top of this thread) which has network monitoring.

_________________
Steve Shipway
UNIX Systems, ITSS, University of Auckland, NZ
Woe unto them that rise up early in the morning... -- Isaiah 5:11