Please Help interpret Sundiag FSTEST failure.
Jeff Vipperman
Hi all,
I can pretty reliably replicate the fstest error in sundiag on /dev/sd2.
I
presume it's the disk and not the controller, since 4 other SCSI devices
on
the controller work fine. I jiggled the cables as well. It looks
to me that this drive is dying, but I wanted some other input. Also,
I reliably have files becoming corrupt in /usr/local, and /export,
which are mounted to this drive, that I have to restore. Particulary
the kernels of the diskless two diskless clients and pine mail program.
I did a format-analyze-test on /sd2 and it didn't come up with any
problems.
I did a /sundiag on /sd2 just after and it passed. This morning the
kernels for the diskless clients were corrupt again and sundiag fails
fstest
again with "compare errors." I'm inclined to replace the drive, but
wanted to make sure I wasn't missing anything...
Thanks for any help in advance!
Jeff
----- sundiag error for /dev/sd2 ----------
999.00.230.7004 05/07/96 07:43:33 pigeon sundiag INFO: *Start all tests*
999.33.999.9014 05/07/96 07:51:27 sd2 ./fstest ERROR: Compare error on
sd2 betwe
en '/usr/local/tmp/tmpdisk1.a00302' and
'/usr/local/tmp/tmpdisk2.a00302', blk 14
7, offset 272.
999.00.230.7012 05/07/96 07:51:34 pigeon sundiag INFO: *Failed test*
(sd2) fstest passes: 0 errors: 1
---------- sundiag error for /kmem when /dev/sd2b was included in
swap---
In addition, the
vmtest failed, until I took /dev/rsd2b swap area out of /etc/fstab!
999.00.230.7004 05/02/96 16:28:30 pigeon sundiag INFO: *Start all tests*
999.04.999.9000 05/02/96 16:29:10 kmem ./vmem ERROR: Test completed.
Found 1 no
ncontiguous miscompare errors.
999.00.230.7012 05/02/96 16:29:11 pigeon sundiag INFO: *Failed test*
(kmem) vmem passes: 0 errors: 1
9
--The machine is a Sun Sparc 2 with two ethernet interfaces; the output
of dmesg:
SunOS Release 4.1.4 (SERVER) #1: Thu Mar 7 07:58:15 EST 1996
Copyright (c) 1983-1993, Sun Microsystems, Inc.
mem = 49152K (0x3000000)
avail mem = 46473216
Ethernet address = 8:0:20:b:6:23
cpu = SUNW,Sun 4/75
zs0 at obio 0xf1000000 pri 12
zs1 at obio 0xf0000000 pri 12
audio0 at obio 0xf7201000 pri 13
sbus0 at SBus slot 0 0x0
dma0 at SBus slot 0 0x400000
esp0 at SBus slot 0 0x800000 pri 3
sd0 at esp0 target 3 lun 0
sd0: <SUN0207 cyl 1254 alt 2 hd 9 sec 36>
sd1 at esp0 target 1 lun 0
sd1: <SUN0207 cyl 1254 alt 2 hd 9 sec 36>
sd2 at esp0 target 2 lun 0
sd2: <Micropolis 1588-15 cyl 1636 alt 3 hd 15 sec 53>
sd3 at esp0 target 0 lun 0
sd3: <NPA31200S cyl 1708 alt 2 hd 15 sec 80>
st0 at esp0 target 4 lun 0
st0: <Archive QIC-150>
st1 at esp0 target 5 lun 0
st1: <Exabyte EXB-4200 4MM Helical Scan>
le0 at SBus slot 0 0xc00000 pri 5
dma1 at SBus slot 1 0x81000
esp1 at SBus slot 1 0x80000 pri 3
sd4 at esp1 target 3 lun 0
sd4: <FALCON MK538FB cyl 1995 alt 2 hd 15 sec 80>
lebuffer0 at SBus slot 1 0x40000
le1 at SBus slot 1 0x60000 pri 5
bwtwo0 at SBus slot 3 0x0 pri 7
fd0 at obio 0xf7200000 pri 11
root on sd0a fstype 4.2
swap on sd0b fstype spec size 12150K
dump on sd0b fstype spec size 12136K
SUNW,lpvi0 at SBus slot 2 0x300000 pri 9
--
Jeffrey Vipperman Internet:
j...@acpub.duke.edu
Dept. of Mech. Engrg.
je...@psych.duke.edu
Duke University, Box 90302 Tel: (919) 660-5434, Fax:
660-8963
Durham, NC 27708-0302 WWW:
http://www.duke.edu/~jsv