Nagios Support Forum

hello,

As we were having some issues with some servers where the nscp service stopped during the backup. Each time it happened there was a log in nsclient.log pointing to committed bytes. So I decided to change my memory checks to only monitor physical memory and not committed bytes for these servers, but now the graphs are no longer working, they look lik the attached image now. The counters for committed bytes are now called Series 3 and 4 and the whole graph is not showing any data anymore
Is there any way to fix this, so I am not loosing my historical data?
If there is no easy fix for this, what can I do to prevent this from happening if I ever would have to do this again for another server?

Tx

Willem

Is there any way to fix this, so I am not loosing my historical data?

No, unless you use a custom plugin/template. The data points are already set and you modified your check after the fact... If you remove the RRDs, they should get recreated and the graphs should show up. You will lose the historical data though. You can save the old RRDs but I don't think you will be able to view them.

If there is no easy fix for this, what can I do to prevent this from happening if I ever would have to do this again for another server?

Do not alter your checks in a way that will change the data points.

Ludmill,

I was thinking maybe the best solution would be

1) Rename the service where I need to do some changes with the renaming tool
2) Deactivate the renamed service
3) Make a new service with the changed attributes

If I would later re-activate the dectivated service, the historical perfdata would not be lost right?

Grtz

Essentially, yes, that will work as long as the hostname nor service config name change for the check/rrd.

Ok, I'll do it like this then. Thread can be closed.