This article is part of a series of posts that deep dive into the new Zerto Virtual Replication 5.0 feature set. To view the other posts check out the index post here.
VMware Host Maintenance Mode has been around since the first version of vSphere. It allows administrators to place certain hosts into a state where they are excluded from cluster HA and DRS calculations. This is extremely important if you have a host that is acting up or needs patches.
VMware also leverages maintenance mode as part of its Update Manager product for applying patches and upgrades to ESXi hosts.
Unfortunately, in all previous version of Zerto, the VRA running on each host would cause the maintenance mode operation to fail, and the host would never enter maintenance mode. The only way around this was to manually shut down your VRA before placing the host into maintenance mode.
While it sounds like more of a formality, this actually creates some interesting problems when automating administrative tasks in VMware.
Starting with version 5.0, Zerto will recognize when you place a host into maintenance mode. Once it has recognized this, it will shut down the VRA that lives on the host and allow the host to go into maintenance mode, just like if Zerto wasn’t installed.
When exiting maintenance mode, Zerto will also automate the powering on of the VRA so that as other VM’s are migrated back to the host they can be protected.
The best way to show how this works is with a video. So this video will walk through what happens with Zerto 4.5 and then show you how the new ZVR 5.0 features help fix the issues.
Thanks for watching!
Does this new feature also automatically clear off the recovery VMs off that VRA before shutting the VRA down? As far as we know this is a requirement in order to put a host in maintenance mode. It’s nice that the power off of the VRA is automated. but in reality it’s just as quick right clicking. the actual moving of the recovery VMs has always been the molasses during host maintenance, especially when you have 16 – 32 hosts in a cluster.
There is logic built in so that they VRA will not shutdown until all other VM’s have been migrated off or powered down. From what I understand we have a “wait loop” that looks to see if there are any powered ON vm’s on the host before shutting down.