This message was deleted.
# harvestera
adamant-kite-43734
11/18/2025, 7:02 PMThis message was deleted.
b
bland-article-62755
11/18/2025, 9:04 PMKinda? It depends on what part is ha
bland-article-62755
11/18/2025, 9:05 PMIf you're talking about service then no. For that , you might need even more than one vm to stay up at the same time for it to be ha.
bland-article-62755
11/18/2025, 9:06 PMTypically, though I think most people assume that HA means it's up all the time which you can't have without live migrations
h
handsome-fish-79554
11/18/2025, 9:15 PMThe definition of HA I'm using is one node has a hardware failure and dies, causing the VMs that were on it to fail and then be restarted on a remaining node
handsome-fish-79554
11/18/2025, 9:18 PMbasically automatic hot spare
b
bland-article-62755
11/18/2025, 9:33 PMrwx is only required for live migrations then. So as long as you're good with forcing a VM reboot to do maintenance on a node you should be fine without it
bland-article-62755
11/18/2025, 9:33 PMJust keep in mind that it might take a few minutes for the cluster to realize a node/vm is down.
bland-article-62755
11/18/2025, 9:34 PMI think there are settings for the tolerances, but I don't remember what they are.
h
handsome-fish-79554
11/18/2025, 9:34 PMyeah that's kind of what I thought, basically I'm using the VMWare definition of HA
handsome-fish-79554
11/18/2025, 9:35 PMas long as I can have VMs automatically restart on available nodes after a node failure that's all I need
b
bland-article-62755
11/18/2025, 9:36 PMSure, but if the VM was in the middle of a block write, you might have some corruption and need a fsck.
h
handsome-fish-79554
11/18/2025, 9:38 PMthey're going to be mostly windows VMs and on Linux I always set my fstab to fsck on startup. I understand that there is possible corruption in the event of a node failure, the goal is that the VMs should in most situations come up correctly
2 Views