[Migration] Disk moved, but network still maxed out

[enoch85]

Trying to move a 6000 GB disk, and it's been running for 1 day now (24 hours). I can see that the disk is moved (space allocated, and `nload` said it transferred more than 9000 GB so far) but it's still running...

The disk are LVM-thin, and Discard=off (not set).

So what actually happens in the background? I mean `nload` says that more than 9000 GB is already moved - did it time out? It's still transferring data as it seems.... CPU dropped at the same time as network got even more busy. I'm confused.

 

[enoch85]

I stopped it now, will try with discard=on instead.

 

[fiona]

Hi,
how much is the VM writing to the disk? Every new write to a block that was already migrated needs to be migrated too of course. After the disk migration, the VM's RAM needs to be transferred. What does the migration task log show you?

 

[enoch85]

Hi,
how much is the VM writing to the disk? Every new write to a block that was already migrated needs to be migrated too of course. After the disk migration, the VM's RAM needs to be transferred. What does the migration task log show you?

The customer decided to reboot the server in the middle of the second attempt, super smart move. Haven't heard anything since.