On 5/18/17 0:36 , Matthew Parsons wrote:
> zone1 is a simple (native) fileserver, serving up a zfs filesystem (let's
> call it "TheFiles") mounted via lofs.
> 
> zone2 is an LX container for running backup software, accessing the same
> filesystem via lofs in read-only mode.
> 
> (It's quite possible that last part is very inadvisable, known to be a Bad
> Idea, etc. This was a quick proof-of-concept, was going to switch the
> backup server to access a snapshot...)
> 
> Anyway, attempting to stop zone1 just halted and never timed out. State is
> now "down".
>  dmesg shows repeated failures of zoneadmd failing:
> 
> "unable to unmount '/zones/UUDI/root/mnt/TheFiles', retrying in 2 seconds...
> ..."unmount of '/zones/UUDI/root/mnt/TheFiles' failed
> ...unable to unmount file systems in zone
> ...unable to destroy zone
> 
> Note that while the process on the backup zone appeared to hiccup a bit,
> it's resumed just fine.
> 
> So 1: Is this somewhat expected behavior, undefined, or a surprise?
> 2: Any recommendations on attempts to get zone back to a bootable "stopped"
> state? Would prefer to not have to reboot host.

When you tried to destroy zone1, did zone2 still have the lofs mount active?

Robert



-------------------------------------------
smartos-discuss
Archives: https://www.listbox.com/member/archive/184463/=now
RSS Feed: https://www.listbox.com/member/archive/rss/184463/25769125-55cfbc00
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=25769125&id_secret=25769125-7688e9fb
Powered by Listbox: http://www.listbox.com

Reply via email to