Ad Widget

Collapse

Defining Host Groups / Maintenance Windows on vmware discovered VMs

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • TheOddPerson
    Junior Member
    • Feb 2024
    • 12

    #1

    Defining Host Groups / Maintenance Windows on vmware discovered VMs


    Hello

    I've done my best to take a look around for a solution and it doesn't seem obvious to me.

    Zabbix 6.4
    I have a vCenter server that is monitoring 3 clusters with 100+ VMs of varying OS / Importance / Maintenance schedule.

    I have manually added my vCenter server with the vmware template. This automatically discovered my ESX Hosts, VMs, and placed the VMs in host groups matching the VM folder they're in. - Great!
    Where possible, these servers have Zabbix agents installed and are reporting their statistics.
    I have additional host groups to categorize hosts by Location, OS, Criticality, and Maintenance Schedule. I realize for the first 3 I could use tags and I'm fine with that.

    TLDR: What is the best way to set (differing) maintenance windows my vmware discovered VMs, since maintenance windows require the use of host groups, and I can't set host groups on vmware discovered hosts?

    Another consideration: We will want to be able to manually place the VMs into maintenance in case we are performing maintenance outside of the regular schedule.

    I have thought of using vmware VM folders for this but I'm already using this to classify VM importance. Also I would need to move the VM to a 'manual maintenance' folder in vCenter and move it back afterwards, which means waiting for a sync from vCenter -> Zabbix. (Default 1h) or manually requesting an update, which is alot of legwork for something I do on a regular basis.

    Any advice is appreciated. Thank you.

  • TheOddPerson
    Junior Member
    • Feb 2024
    • 12

    #2
    Since maintenance schedules can only be set by the host group and the host group on vmware discovered vms can only be set by the vmware folder the vm is in, I've switched to using a tag to identify vm criticality and will use vm folder to identify which maintenance schedule the VM uses.

    When doing manual maintenance outside of normal schedule, the criticality tag should be edited beforehand.
    I've also found that a good way to suppress notifications while doing large maintenance is to go to Alerts->Media Types, and turn off the types of notifications you want suppressed.
    This not only suppresses the notification, but allows you to watch the problem generation in zabbix and the attempt to send a notification (fails) which is a good way to test your triggers and notification logic to ensure it works the way you expect without blowing up your phone or your ITSM system.

    Comment

    Working...