Reboots on clustered nodes in production

Hi,

I’m using openSUSE MicroOS for VMs on Proxmox, with k3s and Elasticsearch running on top.

The setup works well, but I noticed that several MicroOS worker VMs rebooted close together after nightly transactional
updates. The cluster handled it, but for Elasticsearch this is obviously not the ideal maintenance pattern.

I know I could manually assign different maintenance windows to each VM. What I’m curious about is how people here handle
this in production-like clustered setups.

Do you typically:

  • keep automatic updates and stagger reboot windows per node
  • disable automatic reboots and handle them externally/orchestrated
  • use some other pattern entirely

I’d appreciate hearing what has worked well for others.

Thanks.