Forum Discussion

thegabeman's avatar
thegabeman
Novice I
9 months ago
Solved

VMware ESXi, 7.0.3, 23307199 Customer asked to move a VM from non-replicated to a replicated LUN (synchronised replication)

Storage team supplied me with a new LUN and made sure it was on the same array so Storage VMotion would use VAAI and be quick. But it wasn’t While waiting and noticing the Storage VMotion was much slower than expected, I checked on the host if VAAI was working and noticed: `esxcli storage core device vaai status get |grep naa.624a937097ffbb24a2f6483c02ea169b -A 5` `naa.624a937097ffbb24a2f6483c02ea169b` `VAAI Plugin Name:` `ATS Status: supported` `Clone Status: unsupported` `Zero Status: supported` `Delete Status: supported` Moving the 3TB VM took 2 hours in total. Because it was late at night I didn’t investigate further but this morning I wanted to dive into it and noticed: `esxcli storage core device vaai status get -d=naa.624a937097ffbb24a2f6483c02ea169b` `naa.624a937097ffbb24a2f6483c02ea169b` `VAAI Plugin Name:` `ATS Status: supported` `Clone Status: supported` `Zero Status: supported` `Delete Status: supported` How come the clone status is now ‘supported’?

  • Hard to say. This is essentially checked by periodic attempts and then VMware marks the disk as supported or not. Potentially (for whatever reason) an attempt failed, so it reverted to software copy. Then after another check, it succeeded and enabled support
  • codyhosterman Thank you for your reply. Could this have anything to do with the LUN being created just hours before starting to use it? Or the replica not really ready yet? Although as far as I know, these are thin so I’d expect the LUN to be created in seconds and immediately ready also on the replica side. All a bit long shots, but it is the first time I witnessed this and I have had more actions where I thought the copy was much slower then expected, but never looked any further. Are there ESXi log queries I could check? Any event that shows the switch between supported / unsupported?
  • That might be it though—it failed during some kind of replication not ready state