We have a box:
- SuperMicro P6DNE dual 150 PPro, 128MB parity EDO
- a single Adaptec 2940UW (one scsi channel)
- a single Adaptec 3940UW (two more scsi channels)
- a 9GB UW Seagate boot disk
- 5 more 9GB UW Seagate drives (mostly 1 partition/drive)
- 1 more 9GB 80-pin SCA Seagate (with cable converter)
- a scsi Jaz, a scsi Exabyte 8505XLS
- an Adaptec/Cogent 21140 ethernet running various
recent drivers from Becker
which has run nicely for a few years on older kernels
(lately Red Hat 4.0 running 2.0.30 )
However, recent attempts to RAID 0 two of the drives
results in an uptime of only 12-24 hours (vs months before!)
resulting eventually in an "aaaaaiiiieeeeeeeeee" scrolling
error message (system crash).
Let me put in a good word for the MD RAID 0 driver:
we have at least 50 PPro boxes running dual 9GB UW
Seagates and Fujitsu's with no problems whatsoever...
(RH 4.1, kernel 2.0.27 ).
We were trying the latest kernels, (2.1.97 the latest),
and invariably it would boot fine, hence obviously
recognizing the ext2 boot partition. However the other
disks are mounted with a script manually, which would
fail.
Running the script, the error message similar for each
and every drive (and the RAID 0 md0 ) is:
"Couldn't find EXT2 superblock ... trying backup blocks
e2fsck Bad Magic number in Superblock while trying to
open /dev/sdc1
They were originally formatted with
mke2fs -c -b4096 -i4096 /dev/sdc1 -m 1 -v
(or possibly -b8192 -i8192 which used to work but
doesn't seem to lately).
Rebooting with the working kernel (2.0.30) finds no
problems with the superblocks.
The problem restated: modern 2.1.9x kernels do
not recognize older ext2 partitions - possibly related
to -b4096 -i4096 or -b8192 -i8192 formatting.
I'm not on this Kernel maillist, so please cc your
responses (and/or tests for me to perform) to me as
well as the list.
Thanks
Doug Fortune
International Datashare Corp
Calgary
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu