[lxc-users] LXD with ZFS and limiting block I/O resources

[Stéphane Graber]

On Wed, May 25, 2016 at 05:05:03PM +0200, tapczan wrote:
> On 25 May 2016 at 16:59, Stéphane Graber <stgraber at ubuntu.com> wrote:
> > Well, now that's just weird... why would LXD think that /dev/zfs is the
> > backing block for your zpool...
> >
> > Can you paste a "zpool status"?
> 
> # zpool status
>   pool: lxd
>  state: ONLINE
>   scan: none requested
> config:
> 
> NAME        STATE     READ WRITE CKSUM
> lxd         ONLINE       0     0     0
>  zfs       ONLINE       0     0     0


Right, that's the problem right here... ZFS doesn't give us a full path
so we have to make a guess and the logic right now is to first look
directly in /dev, then in /dev/mapper, ...

But because your LV is called "zfs", it matches /dev/zfs and LXD moves
on with that...

We may be able to make it slightly smarter by having it skip character
devices which would probably be enough to fix your case.

> 
> errors: No known data errors
> 
> Hmm, important thing is that this zfs pool is build on top of LVM
> volume (which name is "zfs"):
> 
>   --- Logical volume ---
>   LV Path                /dev/vg0/zfs
>   LV Name                zfs
>   VG Name                vg0
> _______________________________________________
> lxc-users mailing list
> lxc-users at lists.linuxcontainers.org
> http://lists.linuxcontainers.org/listinfo/lxc-users

-- 
Stéphane Graber
Ubuntu developer
http://www.ubuntu.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.linuxcontainers.org/pipermail/lxc-users/attachments/20160525/d683b55a/attachment.sig>