re : rp5470 scsi reset help

[Vincent_5]

Hi Guys,
Can anyone help, I have a rp5470 running with hpux11.11 connected with a DS2300 via A6828A scsi card, lately there is a error on dmesg as below

Jun 5 08:00
...
0009200 78370000 00000000
80080000 91fde288 7f4de000 0000da00
NCR chip register dump for: 0x400200a
00: SCNTL3: 9f SCNTL2: 80 SCNTL1: 00 SCNTL0: da
04: GPREG: 0b SDID: 0d SXFER: 1f SCID: 47
08: SBCL: 00 SSID: 8b SOCL: 00 SFBR: 01
0c: SSTAT2: 0a SSTAT1: 08 SSTAT0: 00 DSTAT: 80
10: DSA: 91fdeb00
14: MBOX1: 00 MBOX0: 00 ISTAT1: 00 ISTAT: 00
1c: TEMP: 20001680
24: DCMDDBC: 0d00038a
28: DNAD: 7f401c78
2c: DSP: 20001698
30: DSPS: 7f401000
34: SCRATCHA: ff030065
38: DCNTL: a1 DWT: 00 DIEN: 7f DMODE: 4c
3c: ADDER: 20402698
40: SIST1: 00 SIST0: 00 SIEN1: 97 SIEN0: 8f
44: GPCNTL: 2f MACNTL: 00 SWIDE: 00 SLPAR: 00
48: RESPID1: 00 RESPID0: 80 STIME1: 00 STIME0: fc
4c: STEST3: 80 STEST2: 00 STEST1: 0c STEST0: 74
50: RESV50: 12 RESV51: c0 SIDL1: 00 SIDL0: 65
54: CCNTL1: 01 CCNTL0: 01 SODL1: 00 SODL0: 00
58: RESV58: 00 RESV59: 00 SBDL1: 00 SBDL0: 80
5c: SCRATCHB: 000b0003
60: SCRATCHC: c0ffffff
64: SCRATCHD: 20001680
68: SCRATCHE: 91fdede4
6c: SCRATCHF: 20000b00
70: SCRATCHG: 9f0d1f05
74: SCRATCHH: 20005650
78: SCRATCHI: 09819f1f
7c: SCRATCHJ: 05005940
bc: SCNTL4: 80
PCI configuration register dump:
Command: 0157
Latency Timer: ff
Cache Line Size: 10

SCSI: Resetting SCSI -- lbolt: 2195852, bus: 4 path: 0/8/0/0

SCSI: Reset detected -- lbolt: 2195852, bus: 4 path: 0/8/0/0

SCSI: Unexpected Disconnect -- lbolt: 2199635, dev: 1f04b000, io_id: 40884d2
SCSI Gross Error on 0/8/0/0:
shadowed SIST0 00 shadowed SIST1 04.
SCSI: isrEscape Controller at 0/8/0/0.

SCSI: -- lbolt: 2199635, bus: 4
lbp->state: 30008
lbp->offset: ffffffff
lbp->nominalOffset: 360
lbp->Cmdindex: 1a
lbp->last_nexus_index: 9
lbp->nexus_index: a
uCmdSent: 19005e40 uNexus_offset: 20005628
last lbp->puStatus [0000000041a15608]:
00030068 00030069 00030079 0003007c
next lbp->puStatus [0000000041a15618]:
00030063 00030074 00030058 0003005b
From most recent interrupt:
ISTAT: 0a, SIST0: 18, SIST1: 00, DSTAT: 80, DSPS: 00000000
lsp: 0x0000000000000000
DSAtbl->host_iocb_index: 1a
DSAtbl->host_iocb_addr: 20005e40
scratch_lsp: 0x0000000000000000
c8xx_iocb [ffffffff91fdeb00]:
1a005e80 ff00005b 200010c0 9f0d1f80
00000003 20005e60 0000000a 20005e68
Pre-DSP script dump [ffffffff91fde338]:
80840000 00000768 745c0200 00000000
888c0000 fffffd48 88880000 fffffca8
Script dump [ffffffff91fde358]:
48000000 00000000 7c5cf000 00000000
7c360000 00000000 745c3000 00000000
NCR chip register dump for: 0x400200a
00: SCNTL3: 9f SCNTL2: 80 SCNTL1: 10 SCNTL0: da
04: GPREG: 0a SDID: 0d SXFER: 1f SCID: 47
08: SBCL: a7 SSID: 8d SOCL: 00 SFBR: 01
0c: SSTAT2: 0b SSTAT1: 07 SSTAT0: 00 DSTAT: 80
10: DSA: 91fdeb00
14: MBOX1: 00 MBOX0: 00 ISTAT1: 00 ISTAT: 08
1c: TEMP: 200012c0
24: DCMDDBC: 0d000d14
28: DNAD: 7e0ad2f0
2c: DSP: 91fde360
30: DSPS: 7e0ad000
34: SCRATCHA: ff000064
38: DCNTL: a1 DWT: 00 DIEN: 7f DMODE: 4c
3c: ADDER: 9108b360
40: SIST1: 00 SIST0: 00 SIEN1: 97 SIEN0: 8f
44: GPCNTL: 2f MACNTL: 00 SWIDE: 00 SLPAR: 00
48: RESPID1: 00 RESPID0: 80 STIME1: 00 STIME0: fc
4c: STEST3: 80 STEST2: 00 STEST1: 0c STEST0: 76
50: RESV50: 00 RESV51: c0 SIDL1: 00 SIDL0: 64
54: CCNTL1: 01 CCNTL0: 01 SODL1: 00 SODL0: 00
58: RESV58: 00 RESV59: 00 SBDL1: 00 SBDL0: 80
5c: SCRATCHB: 000b0013
60: SCRATCHC: c0ffffff
64: SCRATCHD: 200012c0
68: SCRATCHE: 91fdece4
6c: SCRATCHF: 20000b00
70: SCRATCHG: 9f0d1f80
74: SCRATCHH: 20005628
78: SCRATCHI: 09819f1f
7c: SCRATCHJ: 19005e40
bc: SCNTL4: 80
PCI configuration register dump:
Command: 0157
Latency Timer: ff
Cache Line Size: 10

SCSI: Resetting SCSI -- lbolt: 2199735, bus: 4 path: 0/8/0/0

SCSI: Reset detected -- lbolt: 2199735, bus: 4 path: 0/8/0/0

The path is referring to the scsi card if I'm correct. attach is the ioscan output for referrence. Question, if I were to replace the scsi card can I have the steps to bring back my volume groups is from vg01 to vg07 mirror s/w . Please advice???

Regards
Vincent

[Ron Irving]

Hi Vincent!!

The path 0/8/0 is slot 9 in the card cage. Before you do something radical, like replacing the scsi card, try increasing the timeout value using the pvchange command.

#pvchange -t 180 /dev/dsk/c#t#d#

...on all of the devices on that chain.

[Steven E. Protter · Accepted solution]

Shalom Vincent,

Key here is:
lbolt: 2199635

Either you swapped out a hot swap disk or a disk went bad. If the disk is bad, you need to replace it.

http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad.pdf

SEP

[Vincent_5]

Hi Guys,
Thanks for the info, can I know the key that you are referring to is referring to which disk on the DS2300 2199735 and I have check with dd command on all the disk there is no indication of fault . PLease advice

Regards
Vincent

[TTr]

There is more here than just lbolt.

> SCSI: Unexpected Disconnect -- lbolt: 2199635, dev: 1f04b000, io_id: 40884d2

The "dev: 1f04b000" is really the disk that is causing this. The "1f" is hex for 31 which is the major number of /dev/dsk devices. The "04b000" is the minor number for the disk. Run "ll /dev/dsk" and you will see that this is the c4t11d0 disk. Check it out, it is probably failing. It is probably the 12th disk in the DS2300.

[Vincent_5]

Hi,
Thanks for the info, I will try and replace the disk and see if it solved the problem. Thanks

Regards
Vincent

[Torsten.]

SCSI ID 11 is slot 11 in a DS2300.

Slot numbering is 1 ... 14.

slot 1 .. 7 is ID 0 .. 6
slot 8 .. 14 is ID 8 .. 14

[TTr]

> SCSI ID 11 is slot 11 in a DS2300

Torsten you are right. SCSI ID 7 is skipped on the disks, it is used by the controller.
I kept thinking this is a fiber box.

[Vincent_5]

Hi Guys,
Does it means that the harddisk on my DS2300 slot 11 is going to fail, please advice.

Regards
Vincent

[TTr]

What it means is that the error is coming from the disk in slot 11. It could be a timeout error because the disk is extremely busy, it could be an intermittent disk failure which may lead to a permanent failure or it could be a timing issue and you need to install a patch. If you have HP support call in and have it checked out and maybe replace the drive. If not, ensure your data is backed up.