Andy Burns wrote:
OK, now that rawhide is bootable/installable again for me I have tried
dmraid again.
With BIOS in SATA=AHCI mode I booted a rescue CD and did
dd if=/dev/zero of=/dev/sda
dd if=/dev/zero of=/dev/sdb
Then changed BIOS to SATA=RAID mode, rebooted went into the ICH7R
option ROM and created two arrays (a 50GB RAID1 from 50GB of each disk
and a 400GB from the remaining 200GB of each disk)
Booted from rawhide 2005-12-31 with command line
linux dmraid vga=773 debug console=ttyS0,115200 console=tty1
Kernel boots and starts /sbin/loader
I choose language/dhcp/http install
Machine retrieves stage2.img, starts anaconda, starts X11
then immediately backtraces the same way as it did before
http://adslpipe.co.uk/dmraidtrace.jpg
I have a via chipset with 2 80gig disks RAID-1 together.
I too grabbed the lastest bootdisk.img and did a http development
install today but the install went through fine until the final reboot:
Here are some interesting things (bugs) I found out:
1. Whether using the 'dmraid' boot arg or not, I could not visually
tell the difference of what anaconda actually detected in terms of a
physical harddisk while in the partitioning screen. There's really no
confirmation that dmraid is actually playing a part of the partitioning
scheme.
2. There's an option to choose a boot loader method: to install on MBR
or the front of the partition 1. What's interesting is the this screen
shows the device as /dev/mapper/viahhegwhg instead of
/dev/mapper/via_hhegwhg. For some reason the underscore character is
chopped out.
3. I choose for LVM to auto partition my disk and the install went
through normally so I *assumed* that the two drives we're mirror from
the very starting, but on the first reboot, the kernel saw two
conflicting logical volumes, then /dev/sdb took precendence for some
reason and /dev/sda was hung out to dry. I disconnected the /dev/sdb
drive and the system booted fine (granted with no raid).
4. I don't think that 'dmraid' was added to the grub.conf file
automatically once installation finished.
I'm still not sure what I should be excepting but I hope the work
continues in this area.
-eric wood
--
fedora-test-list mailing list
fedora-test-list@xxxxxxxxxx
To unsubscribe:
https://www.redhat.com/mailman/listinfo/fedora-test-list
