5.0.4 Upgrade as SCSI disks configure into kernel, flashes SCSI unrecoverable and hangs

5.0.4 Upgrade as SCSI disks configure into kernel, flashes SCSI unrecoverable and hangs

Post by James Frabon » Sat, 12 Jul 1997 04:00:00



Working through a 5.0.2 Host UPGRADE to 5.0.4 Host towards a 5.0.4
Enterprise Software
Install. On boot with 5.0.4 floppy, including "biosgeom defbootstr" (see
Hardware
Description below), I'm hung up as it assesses the SCSI drives for the
first time:
"Please wait while your SCSI disks are configured into the kernel"

I think I hear the primary drive fire OK during this phase and then the
screen appears to
flash the device message for the secondary drive, (example from messages):

%disk     -             -       -       type=S ha=0 id=5 lun=0 bus=0 ht=hf
%Sdsk     -             -       -       cyls=955 hds=64 secs=32 fts=sd

and then for the briefest to moments I taught I taw a puddy...err the word
pattern of
"unrecoverable error".  The screen instantly redraws the background window
which is
displaying the selector for Upgrade vs Fresh still...but control is never
returned to the
console (15 minutes waited).

This is a functioning SCSI chain now under 5.0.2 host on an IBM Fast/Wide
that has
recently been reconfigured to include the secondary drive, the cd & tape
(all 8-bit
devices/ ie SCSI pin-adapters used) which were inserted into the middle of
the chain,
between the adapater and the primary drive which is at the end of the
ribbon save for the
terminator.  Migration from the original hf adapter included:

        adding the secondary harddrive (this caused some /dev/hd2a verses
        /dev/hd1a non-sense when I finally *'d the original mscsi reference
        to an Sdsk on hf1) but the two partitions mount without error each boot
now.

        adding the scsi tape to hf0, then removing both and adding again just one
again

        adding the cd to hf0, then removing both and adding again just one again

        *'ing out the original mscsi reference in mscsi (shown below)
        #'ing out the hf1 in hf (shown below)

In the middle of booting and re-booting while trouble-shooting the removal
of the
non-existant drive the following was recorded in messages...same drive I
believe the
5.0.4 Upgrade was looking at when it hung:

NOTICE: Sdsk: Unrecoverable error reading SCSI disk 2 dev 1/169 (ha=0 id=5
lun=0
) block=57254
Fri Jul 11 01:17:07 1997

NOTICE: Sdsk: Unrecoverable error reading SCSI disk 2 dev 1/169 (ha=0 id=5
lun=0
) block=57254
Fri Jul 11 01:19:47 1997

NOTICE: Sdsk: Unrecoverable error reading SCSI disk 2 dev 1/168 (ha=0 id=5
lun=0
) block=3762
Fri Jul 11 01:21:45 1997

NOTICE: Sdsk: Unrecoverable error reading SCSI disk 2 dev 1/169 (ha=0 id=5
lun=0
) block=0

WARNING: HTFS: can't read super block on HTFS dev hd (1/169)
Fri Jul 11 01:21:46 1997

NOTICE: Sdsk: Unrecoverable error writing SCSI disk 2 dev 1/169 (ha=0 id=5
lun=0
) block=1
Fri Jul 11 01:21:47 1997

NOTICE: Sdsk: Unrecoverable error reading SCSI disk 2 dev 1/168 (ha=0 id=5
lun=0
) block=0

WARNING: HTFS: can't read super block on HTFS dev hd (1/168)
Fri Jul 11 01:21:48 1997

NOTICE: Sdsk: Unrecoverable error writing SCSI disk 2 dev 1/168 (ha=0 id=5
lun=0
) block=1

On re-boot after those messages, no further errors were reported.

Here a few details:

hf------------------------------------------
hf      Y       1       5       0       0       3540    3547    0       0
#hf     N       1       5       0       0       0       0       0       0

mscsi---------------------------------------
*ha     attach  number  ID      lun     bus
*
hf      Sdsk    0       6       0       0
*hf     Sdsk    1       5       0       0
hf      Sdsk    0       5       0       0
hf      Srom    0       3       0       0
hf      Stp     0       1       0       0

Hardware Description------------------------
This is an IBM PS/2 Model 95 MCA unit, now including the Fast/Wide hf
controller, a 2GB
primary (16bit, ID=6), the original 1GB as a secondary (8bit, ID=5), Sony
SCSI CD (8bit,
ID=3), Tanberg Data TDC 4220 (8bit,ID=1)

32Meg Memory, P90, Original hf has been removed, D-Link Ethernet MCA
currently in that
slot waiting for Enterprise connectivity.  3 DigiBoard MC and 2 DigiBoard
EPCA cards.

defbootstr----------------------------------
Have also been through the message about no root= in the
defbootstr...original included
only hd(40)unix, modified that to DEFBOOTSTR=hd(40)unix swap=hd(41)
dump=hd(41)
root=hd(42)...now more message about no root=

/etc/default/boot------------------ORIGINAL
DEFBOOTSTR=hd(40)unix
AUTOBOOT=YES
FSCKFIX=YES
MULTIUSER=YES
PANICBOOT=NO
MAPKEY=YES
SERIAL8=NO
SLEEPTIME=0
BOOTMNT=RO
TIMEOUT=10

/etc/default/boot------------Modified Lines
DEFBOOTSTR=hd(40)unix swap=hd(41) dump=hd(41) root=hd(42)

hwconfig------------------------------------
name=fpu vec=13 dma=- type=80387
name=serial base=0x3F8 offset=0x7 vec=4 dma=- unit=0 type=Standard nports=1

name=floppy base=0x3F2 offset=0x5 vec=6 dma=2 unit=0 type=135ds36
name=console vec=- dma=- unit=vga type=0 12 screens=68k
name=adapter base=0x3540 offset=0x7 vec=14 dma=- type=hf ha=0 id=14 bus=0
fts=sd
name=adapter base=0x3540 offset=0x7 vec=14 dma=- type=hf ha=0 id=7 bus=1
fts=sd
name=mcxx base=0xF1F0 offset=0x0 vec=- dma=- mem=0x00F80000 ports=16 MC/Xi
V6.0.6
name=mcxx base=0xF2F0 offset=0x0 vec=- dma=- mem=0x00FA0000 ports=16 MC/Xi
V6.0.6
name=mcxx base=0xF4F0 offset=0x0 vec=- dma=- mem=0x00FC0000 ports=16 MC/Xi
V6.0.6
name=epca base=0x208 offset=0x0 vec=- dma=- mem=0x000D0000 ports=32 MC C/X
V6.0.7a
name=epca base=0x228 offset=0x0 vec=- dma=- mem=0x000D8000 ports=32 MC C/X
V6.0.7a
name=tape vec=- dma=- type=S ha=0 id=1 lun=0 bus=0 ht=hf
name=disk vec=- dma=- type=S ha=0 id=6 lun=0 bus=0 ht=hf
name=Sdsk vec=- dma=- cyls=545 hds=128 secs=63 fts=sd
name=disk vec=- dma=- type=S ha=0 id=5 lun=0 bus=0 ht=hf
name=Sdsk vec=- dma=- cyls=955 hds=64 secs=32 fts=sd
name=Stp-0 vec=- dma=- Vendor=TANDBERG Product= TDC 4200

df -v---------------------------------------
Mount Dir  Filesystem              blocks      used      free   %used
/          /dev/root               470978    373738     97240    80%
/stand     /dev/boot                30720     12502     18218    41%
/4         /dev/6                  120000     73858     46142    62%
/5         /dev/7                 1831720    190110   1641610    11%
/0         /dev/0                 1500002    780616    719386    53%
/1         /dev/1                 1500002    982214    517788    66%
/2         /dev/2                  778722    565212    213510    73%

/ /stand /0 /1 & /2 are on primary hd
/4 & /5 are on secondary hd

For upgade purposes, we may need to use the /dev/7 to house parts of the OS


 
 
 

1. System hangs with unrecoverable SCSI bus or device hang.

Hi,

I have a problem relating to my Cyrix 6x86 200+ and
Adaptec 2940 SCSI adaptor.

System Config:
        Cyrix 686 P200+
        AHA 2940 SCSI Adaptor
        48M RAM
        Segate 2GB SCSI Harddisk (scsi id 0)
        SCSI CDROM (scsi id 1)

Linux apache.gih.co.uk 2.0.20 #8 Tue Jul 8 00:02:31 BST 1997 i486

Problem:

        When there is heavy access to my scsi disk,
the system hangs with the following messages.

----
aborting command due to time out (pid 1126) scsi0
channel 0 id0 lun0  0x28 00 00 2a 2b d6 00 00 72 00
(abort) Aborting scb0 TCL 0/0/0
-----

Then the system tries to reset the scsi bus. This
is what it thinks...

---
SCSI host 0 abort (pid 1126) timed out resetting
SCSI bus is being reset for host 0 channel 0

aic7xxxx reset target/channel 0/
0 abort reset scb state 01 Data-In Phase SCSISIGI 0x44

abort scb asserted ATW - host device reset in message buffer.
---

System tried to repeat the reset but with the same results.
Then came the following:

---
SCSI host 0 reset timed out again
probably an unrecoverable SCSI bus or device hang.
---

This problem appeared during fsck when the filesystem
is unclean. It also appeared when X starts, or some
other disk intensive application.

Because it is a 6x86 200+, the bus clock is set to 75MHz.
When I set the bus clock to that of a 166+, which is
66MHz, the system is fine. I was able to use the system
for a long time without problems.

When the system hangs, the red led on the scsi card is
on constantly. I tried to see if the same problem appeared
with NT, well, it did hang for a while with the red LED on,
however, NT seems to be able to get out of it. However,
I was not able to do a lot of testing with NT.

Can someone tell me what the problem might be?

Thank you.

Hansel.

--
Hansel Kwok.                                 Telephone: +44 (0)171 5946309    



2. Linux server to Win hosts on home lan....(Newbie)

3. Unrecoverable error writing SCSI disk

4. compile errors

5. NOTICE: Sdsk: Unrecoverable error reading SCSI disk 1

6. Quotas over NFS mounted filesystem?

7. kernel hang up when close /proc/scsi/scsi file

8. netstat

9. Use of a SCSI Flash disk with a T1 AC200 (or any other rack server) ???

10. cdda2wav hangs SCSI-Bus, wrong SCSI-driver ?

11. SCSI bus hangs on a IN-2000 SCSI

12. RS6000/320: SCSI Pinout/external SCSI disk

13. PROBLEM: drivers/scsi/sd.c - Incorrect Reporting of Blocks and Capacity of Large SCSI Disk Arrays