Problem in creating vg

[Carter Jay]

Hello,

I have a problem when i create a vg on a hp9000 L2000, 8 (18.6 Gb) disks on 2 array controllers, 4 internal disks.
When i launch the command, i have in my /var/adm/syslog/syslog.log file:

SCSI: Parity error -- lbolt: 225424, dev: 1f061000
lbp->state: 4060
lbp->offset: ffffffff
lbp->uPhysScript: fbfff000
From most recent interrupt:
ISTAT: 0a, SIST0: 11, SIST1: 00, DSTAT: 80, DSPS: 0014800
lsp: 21e2600
bp->b_dev: 1f061000
scb->io_id: 60002d0
scb->cdb: 28 00 00 00 00 10 00 00 04 00
lbolt_at_timeout: 0, lbolt_at_start: 0
lsp->state: 5
lbp->owner: 0
scratch_lsp: 21e2600
Script dump [1974000]:
09000800 00148000 e25c0004 fbfff7b8
80080000 fbfff090 80080000 fbfff090

And then:

Pre-DSP script dump [fbfff020]:
00000000 00000000 41010000 fbfff288
78347e00 0000000a 78350800 00000000
Script dump [fbfff040]:
0e000003 fbfff500 e0100004 fbfff7b8
9f0b0000 00000006 0a00000a fbfff508
NCR chip register dump for: 0x802003
00: SCNTL3: 3d SCNTL2: 80 SCNTL1: 10 SCNTL0: d8
04: GPREG: 07 SDID: 01 SXFER: 10 SCID: 47
08: SBCL: a6 SSID: 81 SOCL: 06 SFBR: 00
0c: SSTAT2: 08 SSTAT1: 0e SSTAT0: 21 DSTAT: 80
10: DSA: c0ffffff
14: RESV14: ff RESV15: ff RESV16: ff ISTAT: 28
18: CTEST3: 41 CTEST2: 71 CTEST1: f0 CTEST0: 00
1c: TEMP: fbfffe78
20: CTEST6: 00 CTEST5: 24 CTEST4: 00 DFIFO: 00
24: DCMDDBC: 0e000000
28: DNAD: fbfff504
2c: DSP: fbfff048
30: DSPS: fbfff500
34: SCRATCHA: ff00087e
38: DCNTL: a1 DWT: 00 DIEN: 7f DMODE: 4c
3c: ADDER: fbffe548
40: SIST1: 00 SIST0: 40 SIEN1: 87 SIEN0: 8f
44: GPCNTL: 0f MACNTL: 70 SWIDE: 00 SLPAR: 83
48: RESPID1: 00 RESPID0: 80 STIME1: 00 STIME0: 0c
4c: STEST3: 80 STEST2: 20 STEST1: 0c STEST0: 76
50: RESV50: 80 RESV51: 20 SIDL1: 00 SIDL0: 00
54: RESV54: 80 RESV55: 20 SODL1: 00 SODL0: 7e
58: RESV58: 80 RESV59: 20 SBDL1: 00 SBDL0: 7e
5c: SCRATCHB: 00000000
60: SCRATCHC: 00000000
64: SCRATCHD: 00000000
68: SCRATCHE: 00000000
6c: SCRATCHF: 00000000
70: SCRATCHG: 00000000
74: SCRATCHH: 00000000
78: SCRATCHI: 00000000
7c: SCRATCHJ: 00000000
PCI configuration register dump:
Command: 0157
Latency Timer: ff
Cache Line Size: 10

SCSI: Resetting SCSI -- lbolt: 224123, bus: 6
SCSI: Reset detected -- lbolt: 224123, bus: 6

DIAGNOSTIC SYSTEM WARNING:

Aug 9 13:43:29 manon vmunix: The diagnostic logging facility has started receiving excessive

Aug 9 13:43:29 manon vmunix: errors from the I/O subsystem. I/O error entries will be lost

Aug 9 13:43:29 manon vmunix: until the cause of the excessive I/O logging is corrected.

Aug 9 13:43:29 manon vmunix: If the diaglogd daemon is not active, use the Daemon Startup command

Aug 9 13:43:29 manon vmunix: in stm to start it.

Aug 9 13:43:29 manon vmunix: If the diaglogd daemon is active, use the logtool utility in stm

Aug 9 13:43:29 manon vmunix: to determine which I/O subsystem is logging excessive errors.

Could you help please

Thanks in advance for your help

Cheers

Jerome

[Vincent Farrugia]

Hello,

Ok... SCSI resets.

This could be many things unfortunately. Check these out:

1. Duplicate SCSI IDs
2. Faulty controller
3. Faulty harddisk
4. Bad termination (Check temrinators for bent pins, check SCIS bus layout)
5. Faulty cables (Check cables for bent pins)
6. Faulty HBA.

HTH,
Vince

[Carter Jay]

Hi Vince,

Thanks for your answer

I've already looked at the ID (they are different).
I've swapped the controllers (One of them seems to slowler on each situation).
How can I check for bad disk (without trying to plug them one at the time).
Terminations seems to be good.
Cables too.
In ioscan, i've just got one controller disks (I don't know if there's a master...)

Best Regards


Jerome

[Pete Randall]

Jerome,

You can exercise the disk with a dd command:
dd if=/dev/rdsk/c?t?d? of=/dev/null bs-1024k

If you get errors, you've probably got a bad disk.

Pete

[Vincent Farrugia]

Hello,

Seems like disk c6t10d0 is the culprit, correct me if I'm wrong?

Does this happen when trying to use the other disks?

Vince

[Carlos Fernandez Riera]

Take a look to /dev/rdsk/c6t1d0 disk....



SCSI: Parity error -- lbolt: 225424, dev: 1f061000

[Vincent Farrugia]

Hello, (again)

Actually, SCSI resets is almost never caused by a faulty disk. Even if you do a dd on it and SCSI resets occur, it does not mean specifically that the disk is faulty.

I've seen cases where all disks were giving SCSI resets, but not a single one of them was faulty.

If ONLY the disk is question is giving out SCSI resets, then, yes, probably it's a faulty disk. But check all the other disks with dd, if you see the disk flashing on and off, SCSI resets are occurring on the other disks as well -> this is not a disk issue but a bad bus issue.

That is why I asked whether the same thing happens on the other disks as well.

HTH,
Vince

[Sanjay_6]

Hi Carter,


dev: 1f061000 ---> c6t1d0

61000 --> c6t1d0

Hope this helps.

Regds