Results 1 to 4 of 4

Thread: NSCA monitored host/services - Host flapping

  1. #1
    Member
    Join Date
    May 2013
    Location
    France - Valence
    Posts
    91

    NSCA monitored host/services - Host flapping

    Hi,
    I configured Shinken to monitor hosts through NSCA passive checks, but I have a problem because my hosts status are flapping as soon as a service is not ok ... let me explain :

    My host ('fvc318&#039 is notifying :
    - its own status thourgh an host_check
    - several services status through NSCA

    As soon as all host services are OK, the host status is OK thanks to host_check NSCA notification. But as soon as a service is not OK the host status is automatically changed to DOWN !

    Am I missing some configuration ? Hereunder, an abstract of my logs and my host/service configuration ...

    Many thanks



    Shinken logs :
    [tt]
    [1384519949] PASSIVE SERVICE CHECK: fvc318;nsca_printer;0;Ok - Test de Fred|'Cut Pages'=3082c 'Retracted Pages'=1549c
    [1384519961] PASSIVE SERVICE CHECK: fvc318;nsca_disk;0;OK: All drives within bounds.|'C: %'=95%;96;98 'C:'=383.835G;386.527;394.579;0;402.6 32
    [1384519972] PASSIVE HOST CHECK: fvc318;0;Ok - service 'nscp' running
    [1384519973] HOST ALERT: fvc318;UP;HARD;1;Ok - service 'nscp' running
    [1384519974] HOST NOTIFICATION: mohier;fvc318;UP;notify-host-by-email;Ok - service 'nscp' running
    [1384519974] PASSIVE SERVICE CHECK: fvc318;nsca_cpu;0;OK CPU Load ok.|'5m'=5%;50;80 '1m'=9%;50;80 '30s'=9%;50;80
    [1384519979] PASSIVE SERVICE CHECK: fvc318;nsca_printer;0;Ok - Test de Fred|'Cut Pages'=3083c 'Retracted Pages'=1550c
    [1384520002] PASSIVE HOST CHECK: fvc318;0;Ok - service 'nscp' running
    [1384520006] PASSIVE SERVICE CHECK: fvc318;nsca_memory;1;WARNING: physical memory: Total: 11.9G - Used: 6.52G (54%) - Free: 5.41G (46%) > warning, virtual memory: Total: 8T - Used: 343M (0%) - Free: 8T (100%), paged bytes: Total: 23.9G - Used: 8.13G (34%) - Free: 15.7G (66%), page file: Total: 23.9G - Used: 8.13G (34%) - Free: 15.7G (66%)|'physical memory %'=54%;50;90 'physical memory'=6.52299G;5.96699;10.74;0;11.933 'virtual memory %'=0%;50;90 'virtual memory'=342.633M;4194303.938;7549747.087;0;83 88607.875 'paged bytes %'=34%;50;90 'paged bytes'=8.13299G;11.932;
    [1384520009] HOST ALERT: fvc318;DOWN;HARD;1;CRITICAL: Host is not alive !
    [1384520010] HOST NOTIFICATION: mohier;fvc318;DOWN;notify-host-by-email;CRITICAL: Host is not alive !
    [1384520022] PASSIVE SERVICE CHECK: fvc318;nsca_disk;0;OK: All drives within bounds.|'C: %'=95%;96;98 'C:'=383.835G;386.527;394.579;0;402.6 32
    [1384520032] PASSIVE HOST CHECK: fvc318;0;Ok - service 'nscp' running
    [1384520034] HOST FLAPPING ALERT: fvc318;STARTED; Host appears to have started flapping (53.5% change >= 50.0% threshold)
    [1384520034] HOST ALERT: fvc318;UP;HARD;1;Ok - service 'nscp' running
    [/tt]

    Host configuration :
    [tt]
    define host{
    host_name fvc318
    _ITEMSID 1
    _ITEMTYPE Computer
    alias fvc318 / Computer-1 (Notebook)
    address 192.168.0.100
    parents
    check_command check_nsca_host_alive
    check_interval 60
    retry_interval 1
    max_check_attempts 1
    check_period 24x7
    active_checks_enabled 0
    passive_checks_enabled 1
    check_freshness 1
    freshness_threshold 604800
    realm All
    process_perf_data 1
    notification_interval 120
    contacts mohier
    notification_period Borne totem
    notification_options d,u,r
    }[/tt]

    Services configuration :
    [tt]
    define service{
    use template1-service
    host_name fvc318
    service_description nsca_cpu
    _ITEMSID 1
    _ITEMTYPE service
    check_freshness 1
    freshness_threshold 3600
    check_command nsca_cpu
    contacts mohier
    notification_interval 120
    }
    [/tt]

  2. #2
    Shinken project leader
    Join Date
    May 2011
    Location
    Bordeaux (France)
    Posts
    2,130

    Re: NSCA monitored host/services - Host flapping

    An host check should never be a nsca based one. when service are going critical, a host check command is launched to check this. This behavior is inherited from Nagios in fact. Active checks are not possible for host checks?
    No direct support by personal message. Please open a thread so everyone can see the solution

  3. #3
    Member
    Join Date
    May 2013
    Location
    France - Valence
    Posts
    91

    Re: NSCA monitored host/services - Host flapping

    Thanks for your reply !

    Active checks are not possible for my hosts.

    Host alive checks are based upon an nsca host_check service which is seen as a "PASSIVE HOST CHECK" by Shinken ... I supposed it should be enough for that Shinken considers that my host is alive ? The active checks are not enabled for my host so it does not seem logical to me that Shinken tries to actively check my host
    ... it should consider the last passive host check result.

  4. #4
    Shinken project leader
    Join Date
    May 2011
    Location
    Bordeaux (France)
    Posts
    2,130

    Re: NSCA monitored host/services - Host flapping

    Yep, as we start to talk in a recent github ticket, this behavior is not logic (it's inherited from nagios but seems not good in fact). The good way should be to NOT launch an active check if the host is passive only, and only use it's current state. It's not as easy to implement but should be possible
    No direct support by personal message. Please open a thread so everyone can see the solution

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •