VMware

50/50 On Enabling vSphere HA “VM Monitoring”

As the subject of this post, I’ve never been sure of whether or not I should enable it, something in my bones always feels like something crazy will happen and VM’s will just be restarted during office hours even when they’re actually OK.

Am I being paranoid?

Also is there a flag of some sort set against a VM when a user has actually chosen to shut down a guest gracefully so it doesn’t just start again a few minutes later (I’d always assumed that there was, but don’t know for certain, otherwise how do you perform any guest / VM maintenance?).



View Reddit by Fatboy40View Source

 

To see the full content, share this page by clicking one of the buttons below

Related Articles

2 Comments

  1. It detects failure by [monitoring VMware tools inside the guest](https://docs.vmware.com/en/VMware-vSphere/6.5/com.vmware.vsphere.avail.doc/GUID-62B80D7A-C764-40CB-AE59-752DA6AD78E7.html), and if for some reason the VM heartbeat stops it also monitors for disk and network activity – nothing for two minutes cause a failover. It can also be configured to monitor custom application heartbeats.

    Google “vmware ha deepdive pdf” to find Duncan Eppings excellent free book – I can’t get it to give me a URL on my phone.

    It shows in there (pp30ish) that vCenter marks an HA protected VM as protected or not when it gets started/stopped deliberately. Interestingly though vCenter isn’t involved in the HA restart process, so if the host that disappears is the one with VCSA on HA will still work and restart all the VMs, including VCSA!

  2. Vsphere HA will not reboot a VM unless there is a host failure PSD

    Not having it on during a host failure will just have Powered off VMs sitting on a failed host. As opposed to powered off VMs, you will have VMs rebooted and online

Leave a Reply