Proxmox HA fails after StarWind VSAN node reboot

[roseblack]

Hi everyone,

I’m running a 2-node Proxmox cluster integrated with StarWind VSAN (free version) for shared storage. Everything works fine under normal conditions, but I’ve noticed that when I reboot one of the StarWind nodes, some HA-managed VMs on Proxmox don’t restart properly on the remaining node—even though quorum is maintained.

I’ve already checked that synchronization is healthy before reboot and that both nodes show “synchronized” status. I suspect it might be related to storage timeout or multipath configuration, but I’m not sure where to look exactly.

Has anyone experienced a similar issue with HA failover not triggering correctly after a VSAN node reboot? What specific settings (iSCSI timeout, watchdog, or Proxmox HA config) should I review to make failover more reliable?

[yaroslav (staff)]

Welcome to StarWind Forum!
Do you have the StarWind Plugin installed? If so, please try restarting each node.
Please let me know if your post is referring to Proxmox HA. Please also make sure that the storage is available over iSCSI to both nodes before the restart.
I guess that this issue is more related to the way Proxmox handles failover, rather than the storage configuration. Please note that Proxmox HA configuration is more a subject of Proxmox configuration. I'd suggest asking the Proxmox community https://forum.proxmox.com/. For simplicity, tell them that you use iSCSI-connected shared storage. Make sure to specify the LVM type you use.
Good luck with your project!

[mika5203]

I’ll also review Proxmox HA configuration and LVM type as suggested and follow up on the Proxmox forum if the issue persists.