Ad Widget

Collapse

[ZBX 3.4] Dependencies between Hosts in Templates

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • ColdRaker
    Junior Member
    • Aug 2017
    • 3

    #1

    [ZBX 3.4] Dependencies between Hosts in Templates

    Hello, Cold here,

    I want to implement a dependency between Router and Host. So that if the Router is down, none of the connected Hosts would be displaying an issue.
    The Setup would be:

    Router (down) --> Dont Show Host/VM Triggers
    Router (Up) --> Show Host/VM Triggers

    I know it is possible to manually add this for every Host and make the Host Ping Trigger dependant of the Router Ping Trigger.
    But thats a huge load of work for about 1k Hosts where 3 Hosts are connected to one Router.

    Is there a way around the fact that i cannot make Templates dependant of each other? Also if I were to manually edit the Hosts, after an export all dependencies would be lost.

    But maybe my initial thoughts to accomplish this is wrong I'm also thankfull for any new ideas.
  • kloczek
    Senior Member
    • Jun 2006
    • 1771

    #2
    Theoretically such monitoring may be working but few assumptions must be fulfilled:
    • Monitored gateway will be not used on transporting monitoring data because if it will be down it will be not possible send monitoring data to the srv/prx which could be used when active agent is used or receive polling connections from srv/prx when passive one is used
    • Above means that you must have separated physical paths used on handle monitoring and prod data
    • As long as above is fulfilled you can for example define LLD ilem in which you will be using to populate set of metrics/triggers/etc per monitored host gateway(s)
    • On top of above even network devices (switches and routers) needs to use independent physical paths to monitor them for example over SNMP
    • Even with secondary network used on delivery monitoring data you will need monitoring working from srv and all proxies checking those monitoring paths (at least using ICMP)
    In other words: if your network infrastructure has no physical redundancy on physical network paths you will be not able to implement such monitoring.
    http://uk.linkedin.com/pub/tomasz-k%...zko/6/940/430/
    https://kloczek.wordpress.com/
    zapish - Zabbix API SHell binding https://github.com/kloczek/zapish
    My zabbix templates https://github.com/kloczek/zabbix-templates

    Comment

    Working...