port bandwidth monitors - OK - Current BW in: 0Kbps Out: 0Kb

[nealg]

Here are some samples of the output received. I commented out the community strings with "***". If you need me to send all of the output let me know.

I've attached the output in my next post. I removed the data here because it contained public strings.

[nealg]

Here is all of the output from this command.

[nealg]

I'm thinking this may be a time out issue. Host 10.4.126.2 is a legit host & I see errors in the output for this device. Also I'm seeing some 0's for the perfdata at times for this host. Is the default timeout 2s for these checks? Do you think maybe we should trying changing the timeout to 5s? If yes could you tell me where to look to make this change globally?

[slansing]

Are these devices supposed to be checking over snmp v1? I'd verify that is the right version first. To increase the timeout range for specific devices, you will need to modify the Target line in your mrtg.cfg, or your device's mrtg configuration like so:

Code: Select all

community@router[:[port][:[timeout][:[retries][:[backoff][:[version]]]]][|name]

As an example, to add a timeout of 10 seconds to an existing target line:

Code: Select all

Target[192.168.5.41_13]: 13:[email protected]::10:::2

[nealg]

I've added the following line to my mrtg.cfg file which should change the timeout setting globally to 10s. Let see what happens

SmpOptions: timeout => 10, retries => 2, backoff => 1

[slansing]

That was a type correct? As it should be:

SnmpOptions:

[nealg]

My timeout change doesn't appear to have made any difference. I set the Golbal SNMP timout, ran mrtg command and was able to verify the new timeout setting did take. You can see new output from MRTG command in attachment.

Then ran command from Nagiox XI Gui for SNMP Bandwidth monitor on host sw-internert-01. Still received 0 in/ 0 out. Looked at target config in MRTG host dev file and it is set to use SNMP V2 which is correct.
Target[198.241.2.5_10113]: 10113:[email protected]:::::2

Also I'm not see any errors from the output that I have attached for sw-internet-01 198.241.2.5. When I run the command from the GUI to fetch the perf data the command completes with an OK status.

Back to the drawing board!!! Suggestions?

[scottwilkerson]

Looking at your log, MRTG is running for longer than 10 minutes... That doesn't seem correct, how many ports are you monitoring? Are you getting a ton of timeouts?

This alone could account for the results you are seeing as only MRTG run can happen at a time. If MRTG is locked and 5 minutes later the cron tries to start it again, it will exit if the lock file is present, then the same thing at 10 minutes. this will give you the gap you are seeing.

My only suggestion would be to prune the MRTG config, to eliminate any items you aren't monitoring.

[Box293]

My only suggestion would be to prune the MRTG config, to eliminate any items you aren't monitoring.

Pruning the MRTG config file is important if there are items that are no longer being monitored, especially items which no longer exist on the network.

[nealg]

"BINGO", We have a winner. Removing all of the old MRTG config files seems to have fixed the problem.

I had a lot of old devices which no longer exist in NAGIOS XI but they were still configured for MRTG.

Thanks NAGIOS Team!!!!