On Sat, Jan 15, 2000 at 01:53:23PM -0600, Matthew Hanselman wrote:
> I keep getting "attempt to access beyond end of device" messages in my
> messages log. This has been around in all the 2.2.x kernels I've run, and
> seems to only happen with aic7xxx cards. Since this happens on several
> machines at our site, I am reluctant to dismiss it as a memory/cabling
> problem.
> Device using Wide/Sync transfers at 40.0 MByte/sec, offset 8
I wouldn't be so quick to assume it isn't cabling. We found that
exceeding 20MB/s with (high quality) external cables on single-ended
DPT controllers led to infrequent SCSI errors. Under heavy loads,
command queuing greatly magnified the damage, apparently affecting inodes
the most.
> Jan 14 12:53:55 chewbacca kernel: attempt to access beyond end of device
> Jan 14 12:53:55 chewbacca kernel: 03:40: rw=0, want=606782, limit=593410
> Jan 14 12:53:55 chewbacca kernel: dev 03:40 blksize=2048 blocknr=303390 sector=1213560 size=2048 count=1
The inode damage eventually produced errors very much like this.
Mark.
-
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 - 21:00:13 EST