perfSONAR User Q&A and Other Discussion

[Edward Toth <>]

Hello,

I recently set up some perfSONAR hosts in our network and so far 
everything is working very well.
To be automatically notified if there is a serious drop in throughput or 
an exceptionally high latency I wanted to create some passive Icinga 
checks and send the information to the Icinga host. Before writing the 
checks I wanted to try the nagios checks running on our central 
management node manually to see how they work and what the output is. 
However when I tried the throughput nagios check, I always get the 
following message:

'PS_CHECK_THROUGHPUT UNKNOWN - Unable to find any tests with data in the 
given time range where source...'

The command was: '/usr/lib64/nagios/plugins/check_throughput.pl -u 
https://central.mgmt/esmond/perfsonar/archive -w 0.9: -c 0.5: -r 86400 
-s sonar.node.1 -d sonar.node.2 -p tcp -4'

This seems weird to me, because the Maddash Dashboard does show test 
results. Note that I used different time ranges from 6 hours in the past 
up to several days. It also doesn't matter for which of my perfSONAR 
nodes I run the check.
To see if this was a general problem I also tried 
'/usr/lib64/nagios/plugins/check_owdelay.pl' which returned the expected 
results, again for all of the nodes.

I was wondering if you would know of reason which could explain to me 
why I don't get any throughput information when running the command 
manually.

FYI: We have several Test nodes throughout the network, which were 
manually installed and configured using a mesh config file. They are 
configured in a full mesh, so each node has tests to all the other 
nodes. The test results are sent to a central management node which has 
a maddash server running. I used the documentation on perfsonar.net and 
software.es.net to set everything up.

If you need any more information I will be gad to share it. Also thank 
you for any help you can offer in advance!

Best regards,
Edward Toth