1 |
On 29/01/2019 16:48, Alan Mackenzie wrote: |
2 |
> Hello, All. |
3 |
> |
4 |
> On Tue, Jan 29, 2019 at 09:32:19 -0700, Grant Taylor wrote: |
5 |
>> On 01/29/2019 09:08 AM, Peter Humphrey wrote: |
6 |
>>> I'd rather not have to create an initramfs if I can avoid it. Would it |
7 |
>>> be sensible to start the raid volume by putting an mdadm --assemble |
8 |
>>> command into, say, /etc/local.d/raid.start? The machine doesn't boot |
9 |
>>> from /dev/md0. |
10 |
> |
11 |
>> Drive by comment. |
12 |
> |
13 |
>> I thought there was a kernel option / command line parameter that |
14 |
>> enabled the kernel to automatically assemble arrays as it's |
15 |
>> initializing. Would something like that work for you? |
16 |
> |
17 |
>> I have no idea where that is in the context of what you're working on. |
18 |
> |
19 |
> I use mdadm with a RAID-1 pair of SSDs, without an initramfs (YUCK!). |
20 |
> My root partition is on the RAID. |
21 |
> |
22 |
> For this, the kernel needs to be able to assemble the drives into the |
23 |
> raid at booting up time, and for that you need version 0.90 metadata. |
24 |
> (Or, at least, you did back in 2017.) |
25 |
|
26 |
You still do. 0.9 is deprecated and bit-rotting. If it breaks, nobody |
27 |
is going to fix it!!! |
28 |
> |
29 |
> My command for building my array was: |
30 |
> |
31 |
> # mdadm --create /dev/md2 --level=1 --raid-devices=2 \ |
32 |
> --metadata=0.90 /dev/nvme0n1p2 /dev/nvme1n1p2. |
33 |
> |
34 |
> However, there's another quirk which bit me: something in the Gentoo |
35 |
> installation disk took it upon itself to renumber my /dev/md2 to |
36 |
> /dev/md127. I raised bug #539162 for this, but it was decided not to |
37 |
> fix it. (This was back in February 2015.) |
38 |
> |
39 |
This is nothing to do with gentoo - it's mdadm. And like with sdX for |
40 |
drives, it's explicitly not guaranteed that the number remains |
41 |
consistent. You're supposed to use names now, eg /dev/md/root, or |
42 |
/dev/md/home for example. |
43 |
|
44 |
Cheers, |
45 |
Wol |