Re: Corruption in 2.1.106

Chris Wedgwood (chris@cybernet.co.nz)
Fri, 19 Jun 1998 21:22:17 +1200


On Fri, Jun 19, 1998 at 09:19:46PM +0100, Alan Cox wrote:
> Well in this case I finally found the culprit after the same disk and I/o
> cards ran in another box. Having done the SDRAM shuffle I found a very
> slightly iffy 32Mb SDRAM DIMM which has gone back to the shop down the road
> and been replaced.

The e2fs code is amazingly reliably IMO.

When I read your post and say a few follow ups I was quite worried, mainly
because I'm running 106ac3 on my desktop with about 35GB of disk mostly full
and whilst its not worth backing all of it up (the important stuff is backed
up), it would be a shame to loose it all....

So, I hurriedly dropped to single user mode and forced a fsck on all the
volumes....

Not a bean, nothing. Perfect.

I love this shit, for comparison, I boot NT, do something innocuous like
login, force a chkdsk and reboot (you have to reboot for the chkdsk to work)
and it _always_ finds errors... albeit usually only minor ones.

Another trick I like to play; Get a disk, fill it with files and the zero
10% or so of the drive, choosing where that start at random, etc.

With ext2fsfiles, small size compared to volume). With NTFS, I usually loose the whole
drive, every single damn file.

I definately owe the ext2fs designers a beer or ten. That said, I owe lots
of people beer for all te great kernel and application work that has
occured...

-Chris

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu