Hi, Currently this is the default behavior that the host is rebooted in case of NFS failure.
You can add the line to agent.properties and restart cloudstack-agent to make it effective. reboot.host.and.alert.management.on.heartbeat.timeout=false -Wei On Fri, Mar 28, 2025 at 5:06 AM Antoine Boucher <antoi...@haltondc.com.invalid> wrote: > We experienced unexpected cascading reboots across all hosts, followed by > HA kicking in and migrating VMs. Amid the chaos, we discovered that a newly > added zone-wide NFS server, used only by one stopped test VM, had gone > offline. Once we disabled that NFS server in the UI, everything slowly > stabilized. > > We have a large number of NFS servers online in the zone. Is this expected > behavior? Can one NFS server going offline with just a single stopped VM > trigger mass host reboots? This feels like operational madness. > > Regards, Antoine > > Antoine Boucher > antoi...@haltondc.com > [o] +1-226-505-9734 > www.haltondc.com >
