On 06/24/2009 01:22 AM, David Lehman wrote:
Hi all, I'd like to get a better idea of the perceived usefulness of optionally-partitioned devices. There are two basic examples: - whole-disk filesystem, lvm pv, md component,&c (no partition table) - partitioned md devices We seem to have set out on this road to some extent by accepting patches to use the md subsystem for some fakeraid formats, but I'm wondering how far we want to go. Do we want to support creation of filesystems and other formatting on disks with no disklabel? Do we want to support the creation of disklabels on md devices other than fakeraid arrays? This might have some impact on is the pending UI redesign, as well. Whenever I think about this I end up thinking about having something like "partition table (DOS)" and/or "partition table (GPT)" in the list of available formats for certain device types. With F12 (and RHEL6) approaching fast, it's time to decide what we want to do on this front. Are there more than a handful of users who are interested in this stuff? Is that number going to significantly increase during the lifetime of RHEL6? What are your thoughts?
Partionable mdraid is something which has been discussed before, and for which we have feature requests. So we may want to do that, problem is that we then have to be able to differentiate between a (regular metadata) mdraid set as a disk (iow contains partitions, whole device not all that interesting) and an mdraidset which directly holds a filesystem. mdraid is the only thing besides raw disks where I would consider supporting partitions on. Regards, Hans _______________________________________________ Anaconda-devel-list mailing list Anaconda-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/anaconda-devel-list
