made a test cluster with kvm on libvirt, 3 proxmox nodes, 3 ceph nodes, and one guest to test on. live migration works fine. added the guest to HA under datacenter in the HA tab. verified that its "managed by HA" in the summary and has the option "start at boot". tried a clean shutdown (from the web ui) of the node hosting the guest, and it just sat there still in the same shutdown node. virsh verified that that node really was off. waited a few min, no auto migration. restarted that node, it came back up. pvecm nodes listed it with status 'M', and that guest set to auto start still did not auto start.
am i missing something?
what i did not do was any fencing configuration as i wanted to verify that this stuff worked first, and all the proxmox and ceph nodes are themselves virtual machines on the same host, sharing the same ssd. could this be relevant?
am i missing something?
what i did not do was any fencing configuration as i wanted to verify that this stuff worked first, and all the proxmox and ceph nodes are themselves virtual machines on the same host, sharing the same ssd. could this be relevant?
