Some how nando.net chewed up the following message, which I sent about 4
days ago, so if you got this already, I appologize.

--

This just happened.  No catastrophic crash, no warning.  These are
definitely different blocks than the last time.  (although they all are
around the 13-1400000 range, as before).  Can we please make this stop
happening?  I'll do anything...well, almost.  I just formatted this drive
on saturday, and it was perfect after more than 8 hours of sequential and
random testing.

Hauke's idea sounds quite plausible to me...

What can I do to help???

BTW, I'm running Generic_67.  This is the ncrscsi driver.

-c

root@ros% sh /etc/daily
...Checking setuid files and devices:
Setuid/device find errors:
find: /usr/local/etc/httpd/htdocs/guitar/4/4_non_blondes: Input/output error
find: /usr/local/etc/httpd/htdocs/guitar/acoustic/b/bo_parker: Input/output
error
find: /usr/local/etc/httpd/htdocs/guitar/acoustic/p/paul_simon:
Input/output error
find: /usr/local/etc/httpd/htdocs/guitar/b/baez_joan: Input/output error
find: /usr/local/etc/httpd/htdocs/guitar/b/born_against: Input/output error
find: /usr/local/etc/httpd/htdocs/guitar/c/cooper_alice: Input/output error
find: /usr/local/etc/httpd/htdocs/guitar/d/drivin_and_cryin: Input/output error
find: /usr/local/etc/httpd/htdocs/guitar/f/fields_of_the_nephilim:
Input/output error
find: /usr/share/games/larn: Input/output error
find: /var/spool/ftp/etc: Input/output error
find: /proj/perl5.003/lib/Devel: Input/output error
...
root@ros% fsck -n
** /dev/rsd0a (NO WRITE)
** Last Mounted on /
** Root file system
** Phase 1 - Check Blocks and Sizes

CANNOT READ: BLK 1351872
CONTINUE? yes

THE FOLLOWING DISK SECTORS COULD NOT BE READ: 1351885,
-746216 BAD I=158454
2147483647 BAD I=158454
-196063232 BAD I=158454
INCORRECT BLOCK COUNT I=158454 (4 should be 48)
CORRECT? no

** Phase 2 - Check Pathnames
DUP/BAD  I=158454
CANNOT READ: BLK 1351872
CONTINUE? yes

THE FOLLOWING DISK SECTORS COULD NOT BE READ: 1351885,
 OWNER=root MODE=100555
SIZE=120569 MTIME=Jan 28 07:36 2011

CANNOT READ: BLK 1351872
CONTINUE? yes

THE FOLLOWING DISK SECTORS COULD NOT BE READ: 1351885,
FILE=/usr/local/etc/httpd/htdocs/guitar/f/fields_of_the_nephilim/fields_of_t
he_nephilim.tab

REMOVE? no

** Phase 3 - Check Connectivity
** Phase 4 - Check Reference Counts
** Phase 5 - Check Cyl groups
FREE BLK COUNT(S) WRONG IN SUPERBLK
SALVAGE? no

BLK(S) MISSING IN BIT MAPS
SALVAGE? no

SUMMARY INFORMATION BAD
SALVAGE? no

27577 files, 214289 used, 446321 free (2257 frags, 55508 blocks, 0.3%
fragmentation)
root@ros% more /var/log/messages
...
Sep 10 00:29:23 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 77 77 09 bb 02 00 00 05 00 00
Sep 10 00:29:32 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 6d b6 db 6d b6 db 6d b6 db 6d
Sep 10 00:29:53 ros last message repeated 3 times
Sep 10 00:30:14 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 77 77 09 bb 02 00 00 05 00 00
Sep 10 00:30:37 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 77 77 09 bb 02 00 00 05 00 00
Sep 10 00:30:54 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 00 00 40 00 00 00 00 00 00 00
Sep 10 00:33:45 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 6d b6 db 6d b6 db 6d b6 db 6d
Sep 10 00:34:10 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 77 77 59 da 02 00 00 00 00 00
Sep 10 00:34:24 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 00 00 40 00 00 00 00 00 00 00
Sep 10 00:36:38 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 00 00 40 00 00 00 00 00 00 00
Sep 10 00:46:23 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 77 77 09 bb 02 00 00 05 00 00
Sep 10 00:46:27 ros last message repeated 3 times
Sep 10 00:46:31 ros /netbsd: sd0(ncrscsi0:0:0): medium error, info =
1416959 (decimal), data
 = 00 00 40 00 00 00 00 00 00 00
Sep 10 00:46:33 ros last message repeated 2 times
cmason@ros% uptime
12:49AM  up 1 day, 21:14, 2 users, load averages: 0.63, 0.67, 0.71
         ^^^^^^^^

ARGH!!!!!!!!!

 _____________________________________________________________________
|Christopher Mason - cmason@nyx.net  cmason@nando.net  cmason@cmu.edu |
|"You can always count on a murderer for a fancy prose style."-Nabokov|
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~