General question regarding Nagios.
I know that it is possible to acknowledge an event or alarm in Nagios. But is it possible to clear it?
-
2If by "clear", you mean "make it stop showing up as an error, even though it still qualifies as one", the answer is no, as trikelef says. What you can do is (eg) use a browser plugin like NAGIOS Checker, which can be configured to ignore ACKNOWLEDGED errors (along with many other possible configs).– MadHatterDec 11, 2013 at 9:51
Add a comment
|
3 Answers
There is no concept of "clearing" an alarm with nagios, as one knows from older, physical alerting systems.
There is simply no need. To stay with the traditional concepts; Nagios will clear the alarm by itself when the error has been fixed by you. Until then, you do acknowledge the alarm, so it's silenced - this is an important difference from fixed.
When fixed (where you would clear with other tools) what you want to do with Nagios is to manually trigger an immediate check, this will then clear the alarm.
All that remains is a log entry.
As long as the metric is larger that warning or error threshold it will not stop being a problem.
What you can do is raise the warning and error thresholds but are you sure this is what you want?
There is no "clear" method of doing this, but if it's an error you want to ignore (and you don't have direct access the Nagios' configuration files to delete/modify metrics), you can:
1) Disable active checking for the service in question,
2) Issue a passive check from the service screen and report it "OK". This will cause it to clear the error.
With active checking disabled, it will no longer be updated. If you're using passive checks too, you can make Nagios ignore any further passive checks, too.
This is just a workaround and is not really the designed way to deal with alerts, but it works.
