Francesco Talamona wrote:
> On Saturday 19 April 2008, Roy Wright wrote:
>> Looking thru dmesg and /var/log/messages, it looks like there are no
>> attempts to start the array until I manually try.
>>
>> Any hints on what I'm missing?
>
> Personal experience:
>
> 1) don't mix raidtools stuff with mdadm, use only the latter (I'm not
> saying you used raidtools, but I found a lot of misleading
> documentation lying around)
>
> 2) check carefully the UUID of *all* the partitions, I had exactly the
> same issue, that I discovered to be caused by a leftover partition that
> was part of a different raid set (spurious UUID). At some point in the
> bootup the correct set were disassembled.
>
> 3) evms can badly intefere with mdadm (or it was LVM?): try to modify a
> partition/raid setup and it always appears busy, preventing any
> editing.
>
Thank you.
I only used mdadm following the gentoo.org docs and gentoo-wiki howtos.
Just confirmed all three drives and the mdadm.conf UUID's are the same.
evms not installed.
lvm2 not installed. yet.
I think I found the answer in the man page:
--auto-detect
Request that the kernel starts any auto-detected arrays.
This can only work if md is compiled into the kernel
-- not if it is a module. Arrays can be auto-detected
by the kernel if all the components are in primary MS-DOS
partitions with partition type FD. In-kernel autodetect
is not recommended for new installations. Using mdadm to
detect and assemble arrays -- possibly in an initrd -- is
substantially more flexible and should be preferred.
Basically it seems that I need to add a "mdadm --assemble --scan" to a
startup file.
Thanks again!
Have fun,
Roy
--
[email protected] mailing list
