Page 1 of 1

SNMP Plugin or Script to get I/O wait for host

Posted: Mon Mar 14, 2016 10:15 am
by brdr
Hi folks,

We have XI5.5.2.

I've been looking for a nagios plugin (perhaps a script that calls snmp) that will retrieve the 'I/O Wait', as seen from MPSTAT or VMSTAT commands, for a remote Linux host using snmp commands. This plugin would also need to -w and -c thresholds.

Do you by chance have/or know of plugin?

Thx

Re: SNMP Plugin or Script to get I/O wait for host

Posted: Mon Mar 14, 2016 12:34 pm
by gormank
Have a look in CCM under configuration wizards, at the snmpwalk wizard. It will walk your host and provide a list of OIDs you can monitor. If you find what you need, you can save it and edit to make it generic.

Re: SNMP Plugin or Script to get I/O wait for host

Posted: Mon Mar 14, 2016 2:15 pm
by brdr
It is my understanding that 'CPU IO Wait' percentage does not have a OID attached to it as it is calculated internally by OS.

With that said, since we have the other percentages once could write plug-in or script to make the calls for the percentages and then do so simple math to derive IO Wait. Just thought maybe it was done in the past :-)

CPU
percentage of user CPU time: .1.3.6.1.4.1.2021.11.9.0
raw user cpu time: .1.3.6.1.4.1.2021.11.50.0
percentages of system CPU time: .1.3.6.1.4.1.2021.11.10.0
raw system cpu time: .1.3.6.1.4.1.2021.11.52.0
percentages of idle CPU time: .1.3.6.1.4.1.2021.11.11.0
raw idle cpu time: .1.3.6.1.4.1.2021.11.53.0
raw nice cpu time: .1.3.6.1.4.1.2021.11.51.0

Re: SNMP Plugin or Script to get I/O wait for host

Posted: Mon Mar 14, 2016 2:21 pm
by gormank
Sorry, I was thinking it had an OID...

Re: SNMP Plugin or Script to get I/O wait for host

Posted: Mon Mar 14, 2016 4:18 pm
by tgriep
I did a quick search and didn't find a plugin for this but I thing I found the OID that may be what you are looking for.

Code: Select all

1.3.6.1.4.1.2021.11.54
The number of 'ticks' (typically 1/100s) spent
waiting for IO.

This object will not be implemented on hosts where
the underlying operating system does not measure
this particular CPU metric. This time may also be
included within the 'ssCpuRawSystem(52)' counter.

On a multi-processor system, the 'ssCpuRaw*'
counters are cumulative over all CPUs, so their
sum will typically be N*100 (for N processors).