| View previous topic :: View next topic |
| Author |
Message |
Atreyou40
n00b
Joined: 21 Jun 2006
Posts: 7
|
 Posted: Thu Jun 29, 2006 4:06 pm Post subject: Challenges SA 5i driver after migration |
 |
|
Well,
I just migrated from a 2.4 kernel to a 2.6 kernel (running on a DL380), and I'm having a couple issues:
Can't mount my second logical drive
Can't get eth0 to load with its config
Problem #1
google has many many threads dedicated to the challenges of booting the / partition using this filesystem, and I've figured something important out. With a 2.4 kernel, the appropriate fstab and lilo.conf syntax are:
/dev/cciss/discX/partY => where X = the logical drive, and Y = the partition on that drive
When running a 2.6 kernel, the appropriate syntax is:
/dev/cciss/cAdBpC => where A = the scsi controller, B = the logical drive, and C = the partition
Since I'm already migrated, I'm not really concerned about the 2.4 syntax. What I am concerned about is that my kernel can mount /dev/cciss/c0d0, but canNOT mount /dev/cciss/c0d1. They are referenced exactly the same in /etc/fstab:
| Code: |
/dev/cciss/c0d0p3 / ext3 noatime 0 1
/dev/cciss/c0d0p2 none swap sw 0 0
/dev/cciss/c0d0p1 /boot ext2 noauto,noatime 1 3
/dev/cciss/c0d1p1 /tools ext3 noatime 0 2
/dev/fd0 /mnt/floppy auto noatime 0 0
/dev/cdrom /mnt/cdrom iso9660 noatime,ro 0 0
|
I have absolutely NO PROBLEM mounting that drive once the system is booted using the following command:
| Code: |
mount /dev/cciss/c0d1p1 /tools
|
and a quick view of /etc/mtab shows the partition is mounted, I can do an ls, and run things from it - it's great.
The boot message I get is "failed to mount, could have bad superblock" or something, but it checks out ok. I'm sure that I'm just doing something retarded, I just don't know what it is.
For more background on some of the cciss issues you could go through this thread: https://forums.gentoo.org/viewtopic-t-10380-highlight-dev+cciss+c0d0.html or just search /dev/cciss in google or the forum engine. Many people have had boot up issues, but I can't find anything on this.
Problem #2:
I installed from a custom stage4. After changing my baselayout IAW the migration guide, and following post-emerge instructions I assumed that once I symlinked net.lo to net.eth0 my nic would load with its address information. This is NOT the case. I consulted the gentoo handbook: http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=4&chap=1 to verify I was configuring things right. The major difference was my /etc/conf.d/net was using the "iface_eth0=" syntax. The system the stage4 was built from is still running this syntax and running OK. I can manually assign the address to the interface using ifconfig, however, whether I use the "iface_eth0" syntax that was there previously or the "config_eth0" from the handbook the 2.6, when I boot up, there is no address assigned to eth0, and no gateway set. DHCP works with absolutely NO issues.
So lots of words to describe two problems that should be simple:
Can't mount my second logical drive
Can't get eth0 to load with its manual config
As a side note to both of these problems - when I edit /etc/fstab or /etc/conf.d/net for the syntax associated with a particlar kernel, it doesn't matter. Whether I boot from 2.4 or 2.6 on this DL380, I have the same issues.
More potentially important information - I tried to compile the kernel myself using menuconfig (I couldn't use genkernel because my 2.4 kernel running at the time was NOT conifgured with loopback support). That didn't work - for some reason, even when I selected the appropriate scsi driver, I kept booting to the "kernel panic" that said "could not find root partition".
My 2.6 kernel only seemed to work if I copied the .config file from my 2.4 kernel to the 2.6 /usr/src directory, and loaded that file into menuconfig, saved it, then compiled. I can't explain why - I checked the options I selected for the array controller against what was configured for 2.4, and it was the same. But if a kernel I compile does not find the root partition, it's a MAJOR KING SIZE PITA because of what I first mentioned - that 2.4 uses the discX/partX syntax and 2.6 uses cXdXpX.
Ideas are greatly appreciated. |
|
| Back to top |
|
 |
James Wells
n00b
Joined: 10 Sep 2004
Posts: 57
|
| Posted: Thu Jun 29, 2006 4:35 pm Post subject: |
|
|
Greetings,
| Atreyou40 wrote: | | Can't mount my second logical drive |
Right off the bat, I would suggest changing the 6th field of your fstab to disable bad disk check. If it still fails to mount that way, then I would backup the data and perform a full FSCK, and double check your partition table.
| Quote: | | Can't get eth0 to load with its manual config |
Do you have netplug emerged by chance? I have noticed on my older DL380's that netplugd can take up to 5 minutes before it realizes the netlink is up and starts using the network. |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
Kernel & Hardware
