Hi,
Not satisfied with the way that anaconda orders my partitions when I let it
handle the partition tables, I tried using sfdisk in my %pre section, like so:
sfdisk -uC /dev/sda << EOF
1,26,83,*
27,131,82
158,,fd
EOF
sfdisk -uC /dev/sdb << EOF
1,26,83,*
27,131,82
158,,fd
EOF
And then did the appropriate things in my partitioning section, like:
part raid.1 --onpart=sda3
part raid.2 --onpart=sdb3
raid / --fstype=ext3 --level=1 --device=md0 raid.1 raid.2
Anaconda would then complain about "error opening 'md0': no such device or
address". So I thought that perhaps md was getting started before my %pre
section, and maybe I had to restart it or something, so I went with:
raidstop /dev/md0
echo *** before partition - /proc/mdstat shows:
cat /proc/mdstat
sleep 5
sfdisk -uC /dev/sda << EOF
1,26,83,*
27,131,82
158,,fd
EOF
sfdisk -uC /dev/sdb << EOF
1,26,83,*
27,131,82
158,,fd
EOF
cat >/tmp/raidtab << EOF
raiddev /dev/md0
raid-level 1
nr-raid-disks 2
chunk-size 64
persistent-superblock 1
nr-spare-disks 0
device /dev/sda3
raid-disk 0
device /dev/sdb3
raid-disk 1
EOF
mkraid --configfile /tmp/raidtab -R /dev/md0
raidstart /dev/md0
echo *** in %pre - /proc/mdstat shows:
cat /proc/mdstat
sleep 5
Now, /proc/mdstat shows what I would expect in both places, i.e., no devices in
the first one, and a raid 1 device on md0 in the second... but I still get the
same error. It's weird because I can go to a shell and mkfs and mount /dev/md0
(and /tmp/md0) just fine.
So what gives? Any ideas?
Thanks,
Mart
