[Xymon] question on DURATION and DOWNSECS in alerts

[Kevin VerMeer]

I changed the alert.cfg settings to be 15m instead of 15, to see if that made a difference.
It did not appear to make any difference.  The script was still invoked right away.

-----Original Message-----
From: Kevin VerMeer 
Sent: Monday, April 07, 2014 10:23 AM
To: 'J.C. Cleaver'
Cc: xymon at xymon.com
Subject: RE: [Xymon] question on DURATION and DOWNSECS in alerts

We are just starting with our Xymon setup, so I pretty much assume any issues are on our/my end.
We are running 4.3.13.
It does not appear to be waiting 15m before sending.
The default is minutes, correct?  


-----Original Message-----
From: J.C. Cleaver [mailto:cleaver at terabithia.org] 
Sent: Monday, April 07, 2014 10:01 AM
To: Kevin VerMeer
Cc: xymon at xymon.com
Subject: Re: [Xymon] question on DURATION and DOWNSECS in alerts


On Mon, April 7, 2014 5:55 am, Kevin VerMeer wrote:
> I have a question on what I am seeing in some of the alerts being 
> generated.
> I have a xymon alert script set up for connection events.
> The alert.cfg entry is:
> HOST=* SERVICE=conn COLOR=red
>         SCRIPT /usr/local/xymonutil/alertscripts/noconnectivity.sh
> DURATION>15 REPEAT=15 RECOVERED
>
> To me that says the script should only be invoked if connectivity is 
> down for 15 minutes, repeat every 15 if still down, and one final time 
> when connectivity is back up.
>
> Within the script that gets invoked, a message gets creates like this:
> MSG="Xymon is reporting no connectivty to $STATION.\n  Current time:
> $DATE.\n  Number of seconds down: $DOWNSECS. \n  Time down: $TIMEDOWN. \n"
>
> And one MSG that is generated is
> Xymon is reporting no connectivty to sta10143.
>  Current time: 04/06/14 18:00:55.
>  Number of seconds down: 60.
>  Time down: 00:01:00.
>
> I would have expected the $DOWNSECS variable to be the total time 
> down, even including the original 15 minute DURATION.
> Is that thinking correct?  Or does DOWNSECS only include the time down 
> after the DURATION kicks in?
> Or am I off base on something else?
>

Kevin,

We're doing something similar and - AFAIK - $DOWNSECS is indeed the total duration of the incident (although this may be subject to any flap detection that's enabled).

Q's:
What version are you running?
and Does it actually wait 15m before sending the first message?


Regards,
-jc