On 04/04/2010 23:16, Mark Knecht wrote:
First, thanks for sticking with me on this. I really appreciate it.
Second, I apologize for the length of the reply but it's still not
working and I wanted to try and clearly show the steps I've taken.
Maybe you or someone else will see the step I'm missing.
[snip]
I don't know what to try next.
OK, I think I now understand what's happening here. I regret that I did
not recall this earlier but only the original RAID superblock format
(version 0.90.00) is supported for automatic assembly! I have two
servers that are set up in a similar way as your box, and they both use
this format. It's possible that the docs may be out of date but
/usr/src/linux/Documentation/md.txt says:
"When md is compiled into the kernel (not as module), partitions of type
0xfd are scanned and automatically assembled into RAID arrays. This
autodetection may be suppressed with the kernel parameter
"raid=noautodetect". As of kernel 2.6.9, only drives with a type 0
superblock can be autodetected and run at boot time."
Also, look at this:
http://www.mail-archive.com/linux-r...@vger.kernel.org/msg06215.html
To quote Neil Brown:
"v0.90 can be used with 'in kernel autodetect' (i.e. partition type
0xfd). v1 cannot (I consider this an improvement :-)"
Well, I can't say I agree with him there.
Anyway, it seems that you're using the 1.1 superblock format. So, what
options does this leave you with?
a) Backup the root filesystem, and re-create the array with the regular
superblock format. If necessary, coerce mdadm with -e 0 but it should
be a default.
b) Rely on userspace tools to assemble the array. This means either
having the root filesystem off raid, or using an initrd/initramfs
image.
I'd got for the first option as it keeps things simple and the benefits
of the v1 format are nebulous in practical terms.
Cheers,
--Kerin
