You are not logged in.
Hey there,
I already had my share of problems with my raid setup and kernel/udev updates, but this one is new:
I have a raid1 setup with several partitions on the array which get recognized as /dev/md0p1, /dev/md0p2, etc.
My root-device is on /dev/md0p2 by the way.
After the last kernel-update I get dropped in the rescue shell at boot because the message "waiting 10 seconds for device /dev/md0p2" timed out. In the rescue shell I can see, that indeed the array is properly assembled (cat /proc/mdstat) is fine, but somehow the corresponding partitions (md0p1, sda1, sdb2, etc) are not linked in /dev or recognized at all (/proc/partitions).
However one execution of "blkid" seems to trigger some udev-event and afterwards all partitions and links magically appeared.
Has anyone any idea what causes this behaviour and how to fix it? It would be very enlightening to know which part of my system is actually responsible for setting up the partition-links (udev?).
Offline