Archived community.zenoss.org | full text search
Skip navigation
1 2 3 Previous Next 36343 Views 37 Replies Latest reply: Jun 8, 2010 11:52 AM by arrrghhh RSS Go to original post
  • guyverix ZenossMaster 846 posts since
    Jul 10, 2007
    Check and see if you are ending up with defunct processes as well. I had issues with timeouts causing in my case nagios style http page checks to hang. The only fix was to restart zencommand.

    I have not tested it, but maybe run the command on a nonexistent or down box, and see what the output is, and make it something that conforms to what you normally receive when the box is running normally, but with a value of zero.

    "cdillardhsp" wrote:

     

    Yep, restarting the zencommand daemon kicked the graphs off again. I wish I could provide more info but our Zenoss box is offline for now - bad fan and parts are on order. Will dig into this more once the box is running again.

    Thanks Chet,
    Clay


    On Sun, 2008-03-23 at 08:05 -0400, Chet Luther wrote:

     

     

     

     

     


    On Mar 23, 2008, at 6:49 AM, Clayton Dillard wrote:

     

     

    Maybe one of the Zenoss staff could hop in on this one more time. I
    followed some instructions from a Forum post on Zenoss (referenced
    in one of these list mails) and at first my RTA and loss were being
    graphed. Now I have empty graphs. Strange.


    I'd say that's strange too. How long were they getting graphed for?
    Minutes, hours, days? You can find the RRD file that the data points
    are being written to under ZENHOME/perf/Devices/deviceName/. They'll
    be named like dataSource_dataPoint. See when the last time was that
    these files were updated.

    You could restart zencommand (the daemons responsible for executing
    these checks) with more verbose logging enabled. This will allow you
    to better follow what it is executing and recording in the ZENHOME/
    log/zencommand.log file. Do do this run "zencommand stop" then
    "zencommand start -v 10" as the zenoss user.
    _______________________________________________
    zenoss-users mailing list
    zenoss-users@zenoss.org ([email]zenoss-users@zenoss.org[/email])
    http://lists.zenoss.org/mailman/listinfo/zenoss-users






    Clayton Taylor Dillard

    http://hspcd.blogspot.com/

  • cdillardhsp Rank: Green Belt 127 posts since
    Mar 17, 2007
    Ok, so after adding these commands to check ping latency the zencommand
    daemon crashes regularly. Zenoss folks, do you have a method of
    implementing ping latency with graphing that will not crash the
    zencommand daemon?

    Thank you,
    Clay

    Chet Luther wrote:

     

     

    On Mar 23, 2008, at 6:49 AM, Clayton Dillard wrote:

     

     

    Maybe one of the Zenoss staff could hop in on this one more time. I
    followed some instructions from a Forum post on Zenoss (referenced in
    one of these list mails) and at first my RTA and loss were being
    graphed. Now I have empty graphs. Strange.


    I'd say that's strange too. How long were they getting graphed for?
    Minutes, hours, days? You can find the RRD file that the data points
    are being written to under ZENHOME/perf/Devices/deviceName/. They'll
    be named like dataSource_dataPoint. See when the last time was that
    these files were updated.

    You could restart zencommand (the daemons responsible for executing
    these checks) with more verbose logging enabled. This will allow you
    to better follow what it is executing and recording in the
    ZENHOME/log/zencommand.log file. Do do this run "zencommand stop"
    then "zencommand start -v 10" as the zenoss user.
    _______________________________________________
    zenoss-users mailing list
    zenoss-users@zenoss.org
    http://lists.zenoss.org/mailman/listinfo/zenoss-users


    _______________________________________________
    zenoss-users mailing list
    zenoss-users@zenoss.org
    http://lists.zenoss.org/mailman/listinfo/zenoss-users
  • On 03/04/2008, Clayton Dillard <claytondillard@gmail.com ([email]claytondillard@gmail.com[/email])> wrote:

     

     

    Ok, so after adding these commands to check ping latency the zencommand
    daemon crashes regularly. Zenoss folks, do you have a method of
    implementing ping latency with graphing that will not crash the
    zencommand daemon?



    Have you confirmed that it is the check_ping commands causing your problems. I've been running these commands for months now, admittedly only for a few devices, for months now without a problem.

    I think the zencommand log should show the issues, particularly if the daemon is started with debug logging enabled.

    --
    Regards,

    Graham Bloice
  • snarkout Rank: White Belt 85 posts since
    Dec 20, 2007
    You can use something like these (on linux boxen), though they are fairly hackish, and haven't been thought through more than the 2 min it took me to verify they give me valid output:

    /bin/echo -e \|`/bin/ping -c1 -host- | grep icmp_seq | cut -d" " -f7`

    /bin/echo \|loss=`/bin/ping -c1 -host- | grep 'packet loss' | cut -d " " -f 6`
  • arrrghhh Rank: White Belt 13 posts since
    May 21, 2010

    So did anyone figure out how to do this?  I'm trying to get away from NMIS, which is a horrible and old network management system.  However, it makes these beautiful ping graphs on its own (seemingly) so how can Zenoss do this?  I kind of assumed that's what would be in the performance section - on a router, I don't really care that much about the memory and cpu consumption, a graph of the ping response times would be MUCH better...

  • Andrea Consadori ZenossMaster 863 posts since
    Feb 11, 2008

    try this zenpack

     

    docs/DOC-3467

  • Matt Ray Rank: Zen Master 2,484 posts since
    Apr 5, 2008

    Just reposted what we went over in IRC...

     

    Ping Template and Command Data Source Walkthrough

  • arrrghhh Rank: White Belt 13 posts since
    May 21, 2010

    Cool, thanks guys!

     

     

    Update: I ended up going a slightly different route.  I liked the check_ping method, but did not like the fact that it sent out 5 ICMP packets no matter what.  So I followed this tut - docs/DOC-2513 - and we'll see how I like it.  Seems good so far.

     

    I still wish that there was this type of feature built-in to Zenoss... Zenping is already pinging all my devices, can't I use the stats from that to shape latency graphs in this manner?  It would reduce overhead on all systems, because now I'm doing more pings than necessary I believe.  My template is now pinging devices, as is Zenping.  To me, that just seems silly... Hopefully this will be integrated into the system soon!  Thanks for the support!

1 2 3 Previous Next

More Like This

  • Retrieving data ...