I'm running Xymon 4.3.10 I got an ext script that just displays my Solaris Cluster status. It is just an information only test and will always stay green. At any moment we can just look at cluster status in this test. (we have other test for failures)
But as there is no change in color the different statuses are not logged. Is there a way of saving all the statuses that comes through regardless of no color change? I mean every time it runs (def 5min interval) it saves that output and don't just overwrite the previous. Or even better if there is a change in the status output then save the output.
Sample status output: Thursday, 3 April 2014 9:43:56 AM EST
=== Cluster Resources === Resource Name Node Name State Status Message
FTP-Agent-rs pftp01-zc1 Online Online - Service is online. pftp01-zc2 Offline Offline pftp01-lh-rs pftp01-zc1 Online Online - LogicalHostname online. pftp01-zc2 Offline Offline - LogicalHostname offline.
Cheers, Roland
Den 2014-04-03 0:57, Roland Soderstrom skrev:
I'm running Xymon 4.3.10
I got an ext script that just displays my Solaris Cluster status.
It is just an information only test and will always stay green.
At any moment we can just look at cluster status in this test. (we have other test for failures)
But as there is no change in color the different statuses are not logged.
Is there a way of saving all the statuses that comes through regardless of no color change?
Not directly. Xymon is designed to track status changes, not log all of the data that comes in.
The solution I can think of would be to run xymond_capture and dump the output to a file. If you run
xymond_channel --channel=status xymond_capture --hosts="^myhost$" --tests="^clusterstatus$" >/var/log/clusterstatus.log
then it should save that for you.
Regards, Henrik
I would make it more useful by making it change color.
If there is a change from the last time, turn the color to yellow.
I have a problem app that dies several times a day and night. Rather than have the oncall paged day and night to restart it, I have a script that watches and restarts the app. It then marks the xymon test as yellow. It checks that the app restarted. If not it tries again. If it fails to restart that time too, the test is marked red and the oncall will get paged after a suitable delay. Yes, the developers know, and are looking at it, but there’s a lot of other work going on.
I also have scripts that look up the current color of a test, and will escalate, I.E. first time something isn’t right it goes yellow, the next time it goes red. Obviously, this is for something that it is possible for it to self heal. So you may be able to use that to change back to green.
Paul.
From: Xymon [mailto:xymon-bounces at xymon.com] On Behalf Of Henrik Størner Sent: Thursday, April 03, 2014 4:04 AM To: xymon at xymon.com Subject: Re: [Xymon] Saving histlogs data
Den 2014-04-03 0:57, Roland Soderstrom skrev: I’m running Xymon 4.3.10 I got an ext script that just displays my Solaris Cluster status. It is just an information only test and will always stay green. At any moment we can just look at cluster status in this test. (we have other test for failures)
But as there is no change in color the different statuses are not logged. Is there a way of saving all the statuses that comes through regardless of no color change? Not directly. Xymon is designed to track status changes, not log all of the data that comes in. The solution I can think of would be to run xymond_capture and dump the output to a file. If you run xymond_channel --channel=status xymond_capture --hosts="^myhost$" --tests="^clusterstatus$" >/var/log/clusterstatus.log then it should save that for you.
Regards, Henrik
participants (3)
-
henrik＠hswn.dk
-
Paul.Root＠CenturyLink.com
-
Rolands＠logicaltech.com.au