John said ...
> Hello!
>
> I recently replaced my system board. My old system had two onboard ide
> controllers, with a single 20gb ide drive connected to it. It also had an
> adaptec 2940 with a couple of scsi drives attached. The cache memory on
> the system board went bad, forcing me to replace the board.
Danger Will Robinson! Danger!
> With the new system board the filesystems on the ide drive appear
> corrupted! Running 'e2fsck -vnf' on each partition finds lots of
> problems. The scsi drives are fine -- the scsi controller card hasn't
> been changed though.
Since you've already run fsck and allowed it to 'recover' it's really too
late to get the filesystems back. Hope you had good backups before hand!
> The ide drive is a 20gb with three ext2 partitions, which are 2.5gb, 2gb,
> and 5gb and named hda1 hda2, hda5 respectively.
>
> I booted off the slackware-4.0 bootdisk and rescuedisk and then ran
> "e2fsck -vyf /dev/hda1" This command ran forever and reported a large
> number of errors. After doing this, hda1 will crash the system after
> mounting it and doing "ls /mnt". I ran e2fsck -vnf /dev/hda2 and did the
> same for hda5 both hda2 and hda5 report a large number of errors, which I
> have attached I also to this message.
>
> hda2 and hda5 will still mount, and things will work fine, except some
> files will not open and report back "I/O error". Also, some files are
> 450gb and owned by user 1234234 or something crazy like that. I do not
> believe that running 'e2fsck -vyf /dev/hda2' would be wise?? Wait a
> minute, I'll find out for sure...
>
> I grabbed a 6gb ide drive, and put one partition which is hdc1. I then
> did, "dd if=/dev/hda2 of=/dev/hdc1" which duplicated the hda2 partition
> onto the hdc1 one. I then did "e2fsck -vyf /dev/hdc1". Just like with
> hda1, the hdc1 partition became even more trashed than before. I then
> reran the fsck on hdc1 again and again until it reported no errors. I
> mounted the partition and discovered that all the data that is there
> APPEARS to be intact and correct, yet more than half of the data was
> missing. There must be a better way.
>
> My question is how I might best recover from this corruption. I am hoping
> that this is some kind of c/h/s translation issue, or some similar thing
> where I may be able to enter a boot argument and have it recognize the
> other system's data layout. My goal is to read all the data off the drive
> and then reformat it and put my data back on. I have attempted getting
> the linux disk editor working on the bootdisk system, but that option
> seems kind of unrealistic for recovering all 7 or 8gb.
I've seen exactly what you've just described. I updated my BIOS and
suddenly nothing would mount or run correctly. The BIOS was set to
auto-config the drive. It turns out that the BIOS was reporting Standard
CHS configuration before the update and after the update changed itself
to Extended CHS configuration. The difference is that before the CHS was
1024/16/63 (the supposed real geometry of the drive?) and now was
1024/255/63. None of the filesystems on any of the partitions worked.
As a quick fix I manually set the drive geometry back to the old settings
in the BIOS, did a backup, let the BIOS auto-config the drive again and
re-installed everything.
> - John Frear
> john@data-source.com
-- Peter A. Castro (doctor@fruitbat.org) or (pcastro@us.oracle.com)- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sun Jan 23 2000 - 21:00:19 EST