Nagios XI - Performance Graphs

[jacek]

Hello All,

I have a simple problem - I am thinking that the performance graphs are showing wrong values (maybe instead of Mbps in Gbps).
I tested this by sucking up all my incoming WAN (10 Mbps) downloading a few GB's file, and I saw no influence on the graphs.

I simply don't know how to troubleshoot this, I imagine that I should check following things:
- values that my switch is giving through SNMP
- values being saved by MRTG (maybe there is an conversion?)
- values being drawn by the Performance Graphs plugin and also written underneath them (next conversion?)

Can anyone show me step by step how to achieve this? I'm running the 5R1.0 updated from the previous version (both had the same behaviour).

P.S.
I would also like to do one thing - as I know I have an 10/10 Mbps WAN I would like to "scale" the port as it would be 100% (1Gbps ports), is this achievable?

[jdalrymple]

I simply don't know how to troubleshoot this, I imagine that I should check following things:
- values that my switch is giving through SNMP
- values being saved by MRTG (maybe there is an conversion?)
- values being drawn by the Performance Graphs plugin and also written underneath them (next conversion?)

It's probably a lot dumber than you think. Units are arbitrary and labeled in the perfdata. It's the job of the plugin to interpret the numbers it gets and format the output/units accordingly. Check out the performance data on the advanced tab of one of your services, post it here and we'll show you how to interpret it if it doesn't make sense to you.

I would also like to do one thing - as I know I have an 10/10 Mbps WAN I would like to "scale" the port as it would be 100% (1Gbps ports), is this achievable?

Same as above, maybe just a flag you pass into your check, maybe would require some retooling of the plugin.

[jacek]

Thanks for Your reply, I already looked into that and as You wrote, this doesn't make sense to me:

Code: Select all

Performance Data: in=.000524Mb/s;5;9 out=.154186Mb/s;5;9

Here is the effect of issuing the test command:

Code: Select all

COMMAND: /usr/local/nagios/libexec/check_rrdtraf -f /var/lib/mrtg/172.16.11.190_10134.rrd -w 5,5 -c 9,9 -l M
OUTPUT: OK - Current BW in: 0Mbps Out: .15Mbps|in=.000511Mb/s;5;9 out=.154676Mb/s;5;9

[jdalrymple]

'label'=value[UOM];[warn];[crit];[min];[max]

Values are integers, -(hyphen) or .(period)

So:

Code: Select all

in=.000524Mb/s;5;9 out=.154186Mb/s;5;9

in is the label
.00052 is the value
Mb/s is the UOM

The key here is that Nagios knows nothing about units, it's just courteous enough to let you put that UOM there. As such there is no way to tell Nagios that you don't want Mb/s, you want Gb/s. This is entirely the job of the plugin. It's also important to note that you want that unit to remain the same. So some plugins might be smart enough to change the status output (the words that are in the service detail) from Mb/s to Gb/s or vice versa depending on the number of significant digits, but you'd never want that in perfdata. In perfdata if you start switching units your graphs will end up being 6 orders of magnitude different at night when there is 0 traffic as compared to daytime when there are MegaSomething of traffic.

Is it all making sense?

If so, do you understand what you need to do?

Code: Select all

./check_whatever --help

to see if the output of the perfdata can be customized at runtime. If it can't you get to change the plugin to suit your needs.

[jacek]

Ok, but then it means, that the ".154186Mb/s" value could be 154,186 Mb/s or 154186 Mb/s, am I right?
So if Nagios doesn't care about the values and just presents them, then what's the point of the graphs showing values on the vertical axis? (like a level saying 2Mb/s) etc.

[jdalrymple]

that the ".154186Mb/s" value could be 154,186 Mb/s or 154186 Mb/s, am I right?

The hope is that the data that the plugin is returning is correct. If it's not, what's the point of monitoring. As far as nagios goes though, once the plugin has said to Nagios "The interface is operating at 154,186 Mb/s" nagios itself has no way to validate whether that's true or not. Also nagios doesn't know (nor care) about the difference between Mb/s, Gb/s, etc. A properly written plugin will never change the units in perfdata. Changing the units in the status output is fine, perfdata no.

Example:

service detail.jpg

performance data.jpg

performance graph.jpg

All make sense?

[jacek]

Sorry for such a delay in my reply.
So it seems that maybe I mixed up the graph by changing the commands send out for checks.
Even if not, You provided so detailed info that I will be able to diagnose this when I have enough time.

Thank You for Your help!

[tmcdonald]

Would you like us to keep this open or are you going to continue on your own?

[jacek]

You can close the topic.
I've got it working with the help of this topic

[rkennedy]

Glad to hear! I will now close this thread out, but feel free to open another one if you ever need assistance.