rfc: expectations for partitioning, Fedora.next

[Chris Murphy]

On Feb 24, 2014, at 10:34 AM, Bruno Wolff III <bruno at wolff.to> wrote:

> On Sat, Feb 22, 2014 at 18:33:17 -0700,
>  Chris Murphy <lists at colorremedies.com> wrote:
>> 
>> The bootable raid1 case is actually fragile due to the use of mdadm version 0.9 metadata; 
> 
> I believe that version 1.0 is used for boot partitions, not 0.9. That way they still work with bootloaders that don't understand md raid.

Yes I knew that, I don't know why I wrote 0.9. 
https://bugzilla.redhat.com/show_bug.cgi?id=1046725#c16

However, both are in approximately the same position at the end of the physical device, and still "work" with bootloaders that don't treat them as members in an array. But this is not considered a good practice.
https://bugzilla.redhat.com/show_bug.cgi?id=1046725#c9

If using raid, use a bootloader that understands it. I confirmed in the bug that GRUB2 boots from mdadm metadata version 1.2 superblocks just fine. That's the version we should be using.
https://bugzilla.redhat.com/show_bug.cgi?id=1046725#c18


Chris Murphy