Having trouble disabling notifications on *some/most* nsclient services

Steven Stetson smstetson at yahoo.com
Wed Mar 10 17:37:20 CET 2004


Hello,
I am a new nagios user whose goals for nagios
primarily are to monitor Windows boxes (gasp!) using
nsclient especially to gather perfmon counters and
alert on them as needed.  Thanks to the good
documentation I have been able to get my server up and
running and am currently gathering the perfmon stats
as needed on a handful of test boxes.  There are
roughly 15-20 perfmon counters that I am monitoring on
each of these test boxes (depending on what apps are
installed on the monitored box).  The first 3 boxes I
selected to test on were chosen primarily because they
were non-production machines.  The test was
successfull in that I was able to get the permon data,
set thresholds which were followed, send
notifications, etc.  Once I saw that the clients were
reliable, did not hammer the box, etc I installed
nsclient on a few production machines which also went
well.  I was getting many alerts however, and wanted
to disable notifications for the non-production test
group.  So, i duplicated my service definitions and
created some whose notifications_enabled were set to
'0'.  I had success earlier using this with a service
definition I had set up to get the uptime of the
monitored box via nsclient.  When I configured that
uptime service the little icon which represents
disabled notifications was displayed next to the
service description on the 'Service Status Details For
All Hosts' page.  I read the manual again and the faqs
to try to determine why this might occur.  Listed
below are 2 service descriptions.  The top one does
succeed in disabling notifications, while the 2nd does
not.  

define service{
        host_name               cimsrv
        service_description     Uptime
        check_command           check_nt_uptime
        max_check_attempts      5
        normal_check_interval   60
        retry_check_interval    60
        check_period            24x7
        process_perf_data       1
        notification_interval   1440
        notification_period     24x7
        notification_options    w,c,r
        notifications_enabled   0
        contact_groups          systems
        }


define service{
        host_name               cimsrv
        service_description     Processor Percent
Utilization
        check_command          
check_nt_processor_percent_utilization!60!80
        max_check_attempts      5
        normal_check_interval   1
        retry_check_interval    1
        check_period            24x7
        process_perf_data       1
        notification_interval   1440
        notification_period     24x7
        notification_options    w,c,r
        notifications_enabled   0
        contact_groups          systems
        }

I did finally succeed in disabling notifications, but
only through the web tool's 'Disable notifications for
this service' link
(/cmd.cgi?cmd_typ=23&host=playlist07&service=xxx). 
After I did that I couldn't detect that the .cfg file
with the service definition had been changed.  What is
the cgi doing that I can not?

Here are some of the things I tried unsuccessfully or
thought of:

Program-Wide Filter: 
Set enable_notifications directive in nagios.cfg to 0.
Not appropriate since I do need other notifications.

Service and Host Filters: 
Sheduled downtime.  I suppose I could set up a
downtime which would prevent a notification, but the
max is 99 hours 99 minutes.

Flapping. I can't cause it to think it's flapping.

Notification options.  This is what I'm having trouble
with.  I tried changing the options so that on
recoveries were sent out (which I believe would not
happen since no outage had been sent), but it created
a warning in the sanity check.  And I would like to
keep the sanity check clean.  I also thought that
maybe the reason that uptime worked, but the others
didn't was because nsclient doesn't accept warning or
critical arguments for that check.  I tried to call
the CPULOAD check without the warning and critical
arguments but the client expected them.  

Notification Period. I couldn't find a way to create a
time period which meant never.

Contact Filter:
I suppose I could set up a bogus contact definition
and send no_notify services to that address so that I
wouldn't see this problem.

Since I am in a primarily Windows shop (gasp! again) I
haven't been asked to monitor any *nix boxes with
Nagios.  So this could be a nsclient related issue.

Sorry for the long post.  I wanted to present as much
background as possible to preempt 'did you try this'
replies.  I think everything I've tried is listed
here.  If you don't see something here and think it
might help me - thanks in advance.

Steve

__________________________________
Do you Yahoo!?
Yahoo! Search - Find what you’re looking for faster
http://search.yahoo.com


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null





More information about the Users mailing list