Actually quite a few of my 3 node HCI deployments wound up with only the first host showing up in oVirt: Neither the hosts nor the gluster nodes were visible for nodes #2 and #3.
Now that could be because I am too impatient and self-discovery will eventually add them or it could be because I am using sub-part hardware (Atoms and NUCs in my home lab), which causes race conditions with Ansible (yes, that does happen with oVirt). And in those cases I wound up installing the two other nodes as an act of desperation before restarting the full re-install, either (first) without the hosted-engine option (adding it later) or just with the hosted-engine option, as if they were ordinary compute(-only) hosts. It would have the storage volumes change from being described as single to replica (even if on the Gluster level they had always been replicas) and advance the cluster to a perfectly normal 3-node HCI. But I have no way of knowing if the 3-node HCI base configuration had been injected into the management engine's database already at that point. If it wasn't, then indeed a 1-9 single step increase should basically work, but oVirt still balks at the natural progression from pure replica to erasure code volumes, against which there are many objections hard coded into oVirt Ansible and library Python code. I wish I had the time budget and physical hardware to test this, but unfortunately I can only hope someone else has. _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-le...@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/IYOR6KTYMCOL7ADIDCHJSBODMSGZIY6G/
