I have an ESXI 7 vsan cluster with 4 nodes. Everything has been working good, however, I needed to reboot 1 ESXI node, and during startup, it shows vSAN Recovery with a progress bar (see attachment). It came up ok, so is this normal or is there an issue with how my vsan is setup? Or is there a better process for rebooting 1 host than just putting it in maintenance mode and rebooting?
Hello b1izzard
Yes, this is completely normal - vSAN doesn't flush data from cache-tier devices on power-off/reboot (as that would be slow and/or risky) and essentially has to rebuild and check the metadata on the Disk-Group from data on the cache-tier (e.g. PLOG Recovery).
While this is a faster process in 6.7/7.0 GA than it was in older builds, it is now an order of magnitude faster in 7.0 U1 (e.g. seconds not minutes) and likely you won't be staring at this screen for any considerable length of time once on this build.
Bob
Hello b1izzard
Yes, this is completely normal - vSAN doesn't flush data from cache-tier devices on power-off/reboot (as that would be slow and/or risky) and essentially has to rebuild and check the metadata on the Disk-Group from data on the cache-tier (e.g. PLOG Recovery).
While this is a faster process in 6.7/7.0 GA than it was in older builds, it is now an order of magnitude faster in 7.0 U1 (e.g. seconds not minutes) and likely you won't be staring at this screen for any considerable length of time once on this build.
Bob
Thank you Bob for answering my question, and also explaining why this occurs. I will be updating to 7.0 U1 soon and compare the time difference.
Is that your means vsan recovery due booting esxi just check metadata on cache tier and the taking time depend on how much data reside on the cache tier ?
Yes.