> So I'm reading commit dc215b143 and I'm uncertain how much we want to involve > _legacy_detect_multipath in future facing code. It seems to me that the cases > that support multipath have multipathd present, and cases that don't support > it (desktop) don't have it. > > I would like you to consider that detect_multipath() should return None both > for the `multipath.multipath_supported() == False` case along with the > `multipath.multipath_supported() == True` case where we aren't detecting > multipath. > > _legacy_detect_multipath seems to have history in cases predating proper > multipath awareness. > > I haven't pondered the suggested changes to _legacy very much as I'm > considering the possibility that you should delete _legacy_detect_multipath > outright.
When I first read your comment, I thought that removing the ability to detect multipath disks in environments where multipath-tools is not present almost feels like a regression - and that bringing such a change to 24.04 was questionable. OTOH, I just tried a 24.04 desktop install with a multipath disk and ... _legacy_detect_multipath failed to detect it as a multipath device (missing call to partprobe most likely). All of a sudden I feel much less concerned about "breaking" an existing feature. I will resubmit this MP with the requested change. -- https://code.launchpad.net/~ogayot/curtin/+git/curtin/+merge/481223 Your team curtin developers is subscribed to branch curtin:master. -- Mailing list: https://launchpad.net/~curtin-dev Post to : curtin-dev@lists.launchpad.net Unsubscribe : https://launchpad.net/~curtin-dev More help : https://help.launchpad.net/ListHelp
