VMware Cloud Community
sewert1
Enthusiast
Enthusiast

Shutdown development VMs on HA failover to free up resources

Hi,

we have a customer who wants a two node ha cluster but doesn’t want to waste 50% of the resources of the hardware. He plans an equal number of production and development VMs on the cluster. Further more he does not care about the availability of the development VMs.

Is it possible to automatically shutdown the development VMs to free up resources for the failover of the production VMs on the remaining ESX host in case of a host failure?

If this is possible, we can “overprovision” the two ESX servers with the development VMs and still failover the production VMs.

The only solution I can think of at the moment is scripting.

Thanks in advance,

Sascha

0 Kudos
6 Replies
AntonVZhbankov
Immortal
Immortal

Actually not via standard HA functionality. You can not shut down VMs on HA event, but you can choose which VMs are to be restarted if host fails and which not.

You can do what you requested via script. Connect to vCenter, check if all host are alive, and if not - shutdown VMs form specified list.


---

MCSA, MCTS Hyper-V, VCP 3/4, VMware vExpert

http://blog.vadmin.ru

EMCCAe, HPE ASE, MCITP: SA+VA, VCP 3/4/5, VMware vExpert XO (14 stars)
VMUG Russia Leader
http://t.me/beerpanda
sewert1
Enthusiast
Enthusiast

Thanks for your answer.

How long does HA try to restart the VMs if there are insufficient resources? It will take some time to shut down all the development VMs after my script kicks in. I think I'll have to start the "left over" production VMs with the script, too.

Sascha

0 Kudos
admin
Immortal
Immortal

There is no time limit. If there are not enough resources in the cluster, HA will wait until more resources become available and then attempt the restarts.

Elisha

AntonVZhbankov
Immortal
Immortal

Elisha, how often HA will check if resources are available?


---

MCSA, MCTS Hyper-V, VCP 3/4, VMware vExpert

http://blog.vadmin.ru

EMCCAe, HPE ASE, MCITP: SA+VA, VCP 3/4/5, VMware vExpert XO (14 stars)
VMUG Russia Leader
http://t.me/beerpanda
0 Kudos
admin
Immortal
Immortal

HA should discover the resource changes very quickly (it gets notified when they occur, as opposed to polling for them periodically) and then decide if there are enough resources to perform a failover.

0 Kudos
sewert1
Enthusiast
Enthusiast

Thanks for your help. Scripting seems to be the only solution.

0 Kudos