Results 1 to 4 of 4

Thread: Services Checks with different time frequence

  1. #1
    Junior Member
    Join Date
    May 2015
    Posts
    21

    RESOLVED - Services Checks with different time frequence

    Hello everybody,

    My shinken installation works very fine thnaks to you.
    I'm asked to do some tunning ... not useless but not so needed.

    i made a template(s) which is attributed at first to the host and defines every static values which will be the same for all hosts. I hace few of those and only think to change is the time when the host will be checked

    define host{
    name 5min_short #hard state after 5 min
    check_command check_host_alive
    active_checks_enabled 1
    passive_checks_enabled 0
    check_period 24x7
    notification_period 24x7
    check_interval 3
    retry_interval 1
    max_check_attempts 3

    notification_interval 0 # no need to repeat
    process_perf_data 1
    notifications_enabled 1
    notification_options d,r
    flap_detection_enabled 0
    business_impact 1 # servers 1 #switch 2 # critic server 3 # applicatif 4
    register 0
    }


    But, some service checks need to use differents values (frequences in time) exemple windows
    - want to check disk space every 30 minutes, 3 times max, retry interval 5 minutes
    - want to check cpu usage every 60 minutes, 3 times max, retry interval 10 minutes
    - want to check memory usage every 60 minutes, 3 times max, retry interval 30 minutes

    I never use generic-service, and i write this values in a template or directly in a service.

    Will my template 5min_short run over the service check values of time ? and how to do get icmp as 5min_short want and the service checks diferently (10 min or what ever) ?

    thanks
    Last edited by tome; 07-15-2015 at 02:21 PM.

  2. #2
    Junior Member
    Join Date
    Jun 2015
    Posts
    15
    You can define your service and specify the host there, use 5min_short (so this object will inherit all default value) and then redefine the value you want to override:

    The following will override your default config

    Ex:
    Code:
    define service{
       service_description  PingIcmp
       use                5min_short
       host_name          <YOUR_HOSTS_IP>
       check_command      check_icmp
       check_interval      10
       retry_interval       1
       max_check_attempts   3
       notifications_enabled 0
    }
    Last edited by langlais115; 07-13-2015 at 11:50 AM.

  3. #3
    Junior Member
    Join Date
    May 2015
    Posts
    21
    thank you,

    I'll try even better, to make a service template which will be used in the original check and throwaway generic-service.
    but i have a lot of tunning to do cause there a many services to check...and lots of them useless.
    what would be acceptable charge for the SRV CPU and RAM ? CPU 1H, >90% Critical, RAM 5H > 90% Critical? or I can still go further ? exemple CPU 5H > 90% critical, RAM 10H >90% Critical
    How to anticipate broadcast overload on SW interface ? a FAN on core switch?

    thanks for your help.
    Last edited by tome; 07-13-2015 at 07:51 PM.

  4. #4
    Junior Member
    Join Date
    May 2015
    Posts
    21
    done this way:

    1st

    define service {
    name svc_hardstate_15
    register 0
    check_period 24x7
    max_check_attempts 3
    active_checks_enabled 1
    check_interval 5
    passive_checks_enabled 0
    notification_period 24x7
    notification_options c,r
    notification_interval 0
    notifications_enabled 1
    retry_interval 5
    flap_detection_enabled 0
    process_perf_data 1
    business impact 1
    }

    2nd

    service_description Disks
    use svc_hardstate_15
    register 0
    host_name linux-snmp
    check_command check_linux_disks

    _DETAILLEDESC Overall disks usage
    _IMPACT Depends on disks, cause system instability
    _FIXACTIONS Clean the appropriate disks
    }

Posting Permissions

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