Hello, I’m running Elasticsearch on five Windows virtual machines, and applying Windows updates to them is extremely time‑consuming.Each time I have to stop the shard allocation then proceed with the Windows updates, make sure the machine reboots properly and then allow the shard allocation. Have you found a way to automate this process? Any guidance would be greatly appreciated. Thank you!
I have no experience of Elasticsearch on Windows, but I where I work we use Elasticsearch on Linux servers and those servers regularly install updates and reboot themselves automatically. (There’s some safety checks before a reboot such as, does the cluster contain the expected number of nodes, to avoid a non-Green cluster state being made worse.) We do not stop shard allocation before a server reboots. So reading your post I wonder, why do you have to stop the shard allocation?
Shard allocation can be stopped and starting using the Elasticsearch API Cluster-level shard allocation and routing settings | Elasticsearch Reference so you could write a script which does whatever you need it to do including stopping and starting shard allocation.