Xymon Mailing List Archive search

Acknowledge problem

list Henrik Størner
Thu, 23 Jan 2014 12:45:35 +0100
Message-Id: <user-fad7c9a1d691@xymon.invalid>

 

Den 2014-01-23 11:36, Claessens Jurgen skrev: 
Currently I'm
running xymon 4.3.12 on opensuse 12.3 and I've run into a snag. 

I
have had a yellow disk warning, but as most know, time enough to clear
out disk space. However meanwhile the disk space had increased at such a
rate, that it was completely full. 

What's the problem? Well, it
never came out of acknowledgment. If it turns from yellow to red, I
would expect the ack to be removed. 

Did anybody encounter this
issue before? Is what I expect not possible?
This is "works as
designed". 

As far as Xymon is concerned, once an event has been
ack'ed, then it won't care about it until it has returned to green or
the ack expires. Ack'ing an event means "I know about it, and I'll fix
it" - then it is the admin's responsibility to handle it. If it goes
from yellow to red while he is doing that, is not something Xymon
worries about. 

This is because the primary purpose of an ack is to
stop alerts from being triggered - you dont want to be pestered with
alerts while you are working on an issue. Also, if you have e.g. a disk
that is flipping between yellow and red, then you don't want to
constantly re-ack the event just because the color changed. 

You could
argue that an ack should be associated with a severity - so an
ack-on-yellow would not have effect when the status went red; then you
would have to re-ack it. But both acks should remain in effect until the
status goes green. But that is not how the system works right now.


Regards,
Henrik