On Fri, Oct 06, 2000 at 07:01:15PM +0100, Corin Hartland-Swann wrote:
> Yes - but I was curious as to whether this kernel memory bug had been
> fixed yet, or whether it was still happening. I'd like to upgrade from
> 2.2.14 (because of the security hole) but the do_try_to_free_pages() bug
> is IMHO a much greater problem than the hole.
Well, we'll see how it goes. I guess I'm gonna have to live
with the memory issue, because I have needs for the .18 features.
> Do you have physical access to the machine? Maybe you could re-seat the
> cabling and/or swap out the controller.
The drive failed again. Just on some reads (an ftp connection).
It's really shaping up to be the drive. I opened the box and checked
the cables. They looked fine. So I yanked the drive and I'm running on
half a mirror now. I intend to pound the drive in another box.
> Another sensible idea is to run badblocks on the raw device (as long as
> it's not in write mode it shouldn't hurt the active MD partition). Run it
> a few times and see if it provokes and kernel messages.
Yeah, we'll hit it. I also ran the SCSI cards disk verification
utility last night, to no avail.
> What do the boot= and root= lines in your lilo.conf look like? They should
> point to the root MD device, and not to one of the individual disks, eg:
I'm not booting off of the md device. When it gets to the MD
device, it won't start it (autostarting). /etc/raidtab had both disks
in the config.
In addition, when I booted after I removed the allegedly faulty
disk, the md device would not start up because it couldn't find the
second device. I'd think that it would start up in the degraded state,
not fail to start at all. I guess I'm mistaken.
> Under RAID-1 the raw partition contains a valid ext2 partition, /which is
> slightly smaller than the raw parition/. The remainder (at the end) stores
> the RAID superblock.
Ok, so I *should* be able to mount/fsck it. I'm glad to be sure
of that. This actually makes a little more sense. Remember I mentioned
that I tried to mount one of the partitions raw and the kernel spewed
errors? Could be the faulty disk being faulty.
Thanks,
Joel
--
"Hell is oneself, hell is alone, the other figures in it, merely projections."
- T. S. Eliot
http://www.jlbec.org/
[EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to [EMAIL PROTECTED]
