[zfs-discuss] zfs and mdadm

Ovi Chis ovios.storage at gmail.com
Wed Aug 10 08:39:46 EDT 2016

Why use RAID (zpool) over RAID (md)

On Aug 10, 2016 12:26 AM, "Dirk Jakobsmeier via zfs-discuss" <
zfs-discuss at list.zfsonlinux.org> wrote:

> Hello,
> i've seen a misterious behaviour this morning.
> Debian Testing ( , zfs from sources. I'm using zol on
> this system since several months and i only had to compile with every new
> kernel, no problem. As i always use suspend-to-disk i boot only about one
> time in a month. When doing the last time everythink was ok. I got a new
> kernel last week and when booting this morning my zfs was missing, zfs
> import said all the devices where unavailable. Yes, i've forgot to
> recompile zfs. Doing so did not solve the problem, after reboot again
> devices where unavailable. So back to the old kernel -> no success. Then i
> looked for the 2 disks and found a new meta device (md3) used by mdadm. But
> it was not configured and not used. But the physical devices where my 2
> missing zfs disks (partitions). So i feared that the mdadm did rewrite any
> part of the physical disks but tried to mdadm --stop md3, this did work,
> and zpool import -> ok, success, now i could import my old zpool.
> Is it known that mdadm can assemble a metadevice out of two zpool devices?
> --
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at list.zfsonlinux.org
> http://list.zfsonlinux.org/cgi-bin/mailman/listinfo/zfs-discuss
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.zfsonlinux.org/pipermail/zfs-discuss/attachments/20160810/94dee192/attachment.html>

More information about the zfs-discuss mailing list