Announcement

Collapse
No announcement yet.

LAtence détectée avec centreon-engine

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • LAtence détectée avec centreon-engine

    Bonjour,

    Je viens de migrer de nagios à centreon-engine. Je m'attendais à constater un gain de performance, mais pour le moment j'observe surtout une énorme latence.

    Je supervise en majorité en passif via NSCA ( > 200 machines ). J'ai cependant une dizaine d'hôtes en actif.

    Le load-average est proche de 0 pourtant.

    Cette même configuration sous nagios fonctionnait parfaitement. Pouvez-vous m'aider à comprendre ?

  • #2
    Voici la sortie de centenginestats :

    [email protected]:~# centenginestats
    Centreon Engine Statistics Utility 1.3.2


    Copyright 2003-2008 Ethan Galstad
    Copyright 2011-2013 Merethis
    License: GPLv2


    CURRENT STATUS DATA
    ------------------------------------------------------
    Status File: /var/log/centreon-engine/status.dat
    Status File Age: 0d 0h 0m 38s
    Status File Version: (null)


    Program Running Time: 5d 0h 30m 59s
    Centreon Engine PID: 16722
    Used/High/Total Command Buffers: 0 / 1017 / 4096


    Total Services: 4088
    Services Checked: 4088
    Services Scheduled: 62
    Services Actively Checked: 237
    Services Passively Checked: 3851
    Total Service State Change: 0.000 / 54.800 / 1.913 %
    Active Service Latency: 197.109 / 188046.531 / 96083.864 sec
    Active Service Execution Time: 0.048 / 21.822 / 4.175 sec
    Active Service State Change: 0.000 / 54.800 / 4.796 %
    Active Services Last 1/5/15/60 min: 0 / 0 / 0 / 0
    Passive Service Latency: 0.126 / 42.096 / 9.969 sec
    Passive Service State Change: 0.000 / 34.410 / 1.735 %
    Passive Services Last 1/5/15/60 min: 0 / 0 / 0 / 0
    Services Ok/Warn/Unk/Crit: 3741 / 37 / 46 / 264
    Services Flapping: 0
    Services In Downtime: 0


    Total Hosts: 254
    Hosts Checked: 252
    Hosts Scheduled: 13
    Hosts Actively Checked: 21
    Host Passively Checked: 233
    Total Host State Change: 0.000 / 11.580 / 0.374 %
    Active Host Latency: 0.000 / 455.998 / 396.821 sec
    Active Host Execution Time: 0.000 / 10.148 / 3.792 sec
    Active Host State Change: 0.000 / 11.580 / 4.525 %
    Active Hosts Last 1/5/15/60 min: 0 / 0 / 0 / 0
    Passive Host Latency: 0.127 / 28.821 / 10.140 sec
    Passive Host State Change: 0.000 / 0.000 / 0.000 %
    Passive Hosts Last 1/5/15/60 min: 0 / 0 / 0 / 0
    Hosts Up/Down/Unreach: 239 / 15 / 0
    Hosts Flapping: 0
    Hosts In Downtime: 0


    Active Host Checks Last 1/5/15 min: 6 / 26 / 74
    Scheduled: 0 / 0 / 0
    On-demand: 6 / 26 / 74
    Parallel: 0 / 0 / 0
    Serial: 0 / 0 / 0
    Cached: 6 / 26 / 74
    Passive Host Checks Last 1/5/15 min: 0 / 10 / 16
    Active Service Checks Last 1/5/15 min: 0 / 0 / 0
    Scheduled: 0 / 0 / 0
    On-demand: 0 / 0 / 0
    Cached: 0 / 0 / 0
    Passive Service Checks Last 1/5/15 min: 0 / 126 / 194


    External Commands Last 1/5/15 min: 1246 / 5100 / 14971

    Comment


    • #3
      Capture du 2013-11-06 12:22:50.jpg

      Voici les réglages que j'ai (par défaut).

      Je regarde actuellement cette article : http://documentation.centreon.com/do...rformance.html

      Comment


      • #4
        Voici ce que j'en ressort :

        External command optimizations. If you’re processing a lot of external commands (i.e. passive checks in a distributed setup, you’ll probably want to set the command_check_interval variable to -1.

        Use Setpgid

        This option allow to change plugin process group into they own process group id. This option protect Centreon Engine process from child miss used or bug.For example, if we use nagios check_ping, check_dns, check_dig or check_rbl, don’t disable this option, because, these checks can call kill -KILL 0 on timeout (this is a bug from these plugins) and kill the engine if the PGID is the same as the engine.For maximum performance, this option must be disable.
        • 0 = Don’t use setpgid
        • 1 = Use setpgid (default)

        Format use_setpgid=<0/1>
        Example use_setpgid=1
        External Command Buffer Slots

        Format external_command_buffer_slots=<#>
        Example external_command_buffer_slots=512
        Note This is an advanced feature. This option determines how many buffer slots Centreon Engine will reserve for caching external commands that have been read from the external command file by a worker thread, but have not yet been processed by the main thread of the Centreon Engine deamon. Each slot can hold one external command, so this option essentially determines how many commands can be buffered. For installations where you process a large number of passive checks (e.g. distributed setups), you may need to increase this number.


        Mais je ne connais pas la valeur par défaut donc je n'y ai pas touché.

        Comment


        • #5
          peut être une piste ici : http://forum.centreon.com/showthread...lower-than-ndo

          La latence peut elle être induite par le broker ?

          Comment


          • #6
            J'ai l'impression que le freshness est mal interprété. Lorsque je passe en service en état ok qui a un freshness à 360 secondes.

            exemple :

            [1383922600] EXTERNAL COMMAND: PROCESS_SERVICE_CHECK_RESULT;2sf_[PROVECTIO];Apache;0;|
            [1383922604] Warning: The results of service 'Apache' on host '2sf_[PROVECTIO]' are stale by 0d 0h 1m 3s (threshold=0d 0h 5m 20s). I'm forcing an immediate check of the service.

            Comment


            • #7
              https://forge.centreon.com/issues/4987

              Comment

              Working...
              X