On Wed, 23 Feb 2005, Eric S. Johansson wrote: > as I suspected the raid array was not be automatically detected which > means that everything else fails miserably. The system is operational > with all of the modules (tulip, raid1, md, siimage, dm_mod) are all > loaded and the individual components of the raid array are visible. But > the other detection does not happen even though they have the right > partition type (fd). > > so I figured with an entry in /etc/mdadm.conf, checkfs would be able to > start things up. my changes to /etc/mdadm.conf were: > > DEVICE /dev/hde1 /dev/hdg1 > ARRAY /dev/md0 UUID=91287317:6bf81ee1:31701212:3b243352
Im using a raidtab file in /etc. > but there's a problem (no surprise I'm sure). The lvm2 code looks for > logical volumes first and raid arrays second. Which is the complete > opposite of what I need. I seem to remember reading something talking > about this but I can't find it. > > Closer and closer thanks to commentary from folks here. Any ideas on > how to jump the last gap short of of rewriting checkfd? Yeah this is exactly what the problem is - there needs to be a way of detecting the order of the file-systems so stuff gets started in the correct order. I remember that thread and some people hacked into the scripts to fix it but this is a not a good long-term solution. -- -- gentoo-user@gentoo.org mailing list
