Messages in this thread | | | Date | Wed, 29 Jun 2011 11:40:18 +0200 | From | martin f krafft <> | Subject | Re: nested block devices (partitioned RAID with LVM): where Linux sucks ;-) |
| |
also sprach Ulrich Windl <Ulrich.Windl@rz.uni-regensburg.de> [2011.06.29.0914 +0200]: > 1) The original error message of mdadm about a wrong UUID is > completely wrong ("device busy" would have been correct)
Correct. It would be nice if you could file a bug about this in your distro's bug tracker (or Debian's).
> 2) partitions on unassembled raid legs are activated before the > RAID is assembled, effectively preventing a RAID assembly (I could > not find out how to fix/prevent this)
I think you will find that LVM snatched the PV before mdadm had a chance, hence it was busy. This is a common problem with LVM and RAID1, because LVM (also) scans all devices and because RAID1 is merely a mirroring setup, LVM can use either of the components just as well.
The solution? I thought that there was a patch to LVM that prevented it from using RAID members. But if there isn't, then exclude the devices from its scan.
… and before you claim that Linux sucks, consider how a computer should do it differently. I think you will find that every implementation has to make certain hard assumptions (for it lacks human abstraction and combination abilities), and your current griefs make you feel like Linux does it worst of all. Rest assured: it doesn't. It makes assumptions, but you will find that they are quite sensible.
> I haven't studied the block data structures, but obviously the > RAID metadata is not at the start of the devices. If they were, > a partition table would not be found, and the RAID could have been > assembled without a problem.
You are right, the metadata are not at the start. This is by design.
> I'm sending this message to make developers aware of the problem,
(in which case it might be wise to avoid exclaiming things like "linux sucks"…)
> P.S. Novell Support was not able to provide a solution for this problem in time
News at 11…
-- martin | http://madduck.net/ | http://two.sentenc.es/ "no, 'eureka' is greek for 'this bath is too hot.'" -- dr. who spamtraps: madduck.bogus@madduck.net [unhandled content-type:application/pgp-signature] | |