You are not logged in.
When you rebuild your intramfs, does it use your mdadm.conf?
Offline
Yes, it does.The same config can boot with the old init scripts without any issue (although it hangs during shutdown). Maybe another info can add some insight: if I reference the devices by UUID in the fstab file the system boots normally using the onld scripts, but systemd tries to mount the individual disks instead of the array (and fails at doing so).
EDIT: I opened a bug on the tracker: https://bugs.archlinux.org/task/31257?project=1
Last edited by gondsman (2012-08-23 08:59:04)
Offline
I have this problem to (((
I found related bug
https://bugs.archlinux.org/task/31236
Last edited by Zipfer (2012-11-30 10:42:44)
Offline
@Zipfer: what chipset do you have? I can happily report that the problem was solved for fakeRAID on intel motherboards.
Offline
@Zipfer: what chipset do you have? I can happily report that the problem was solved for fakeRAID on intel motherboards.
No, I have AM3 chipset - Gigabyte GA-MA790XT-UD4P
and i can't mount anything except root directory
thanx for help
Offline
In that case your best bet is to remove /home from fstab and mount it with a script during boot. According to the bug report linked before something is moving with dmraid support, but it's not quite there yet.
Offline
Note: After the systemd/udev update on my system, sda1 was renamed sda2 by the system causing a boot failure similar to what was discussed here. If you experience a boot failure after update/install of the new systemd, check the device names on your system. The consistent naming scheme of the new udev (included within systemd) may cause device name changes throughout your system. (eg. sda1->sda2, eth0->enp2so, etc...)
I have 3 Arch servers that have multiple dmraid arrays each. I have purposely not updated those boxes until I get a better handle on how the naming of the arrays will behave under the new scheme.
David C. Rankin, J.D.,P.E.
Offline