SCSI: Reset detected -- dmesg & stm notification

[Ratzie]

We just received an error that a SCSI Reset was detected.
From syslog...
First
Oct 6 19:01:22 server1 vmunix:
Oct 6 19:01:22 server1 vmunix: SCSI: Reset detected -- path: 0/10/0/0
Oct 6 19:01:22 server1 vmunix: SCSI: Reset detected -- path: 0/10/0/1
Oct 6 19:01:22 server1 vmunix: SCSI: -- lbolt: 654310921, bus: 4
Oct 6 19:01:22 server1 vmunix: lbp->state: 30008
Oct 6 19:01:22 server1 vmunix: lbp->offset: ffffffff

Then:
Oct 6 19:03:55 server1 vmunix: SCSI: Reset detected -- path: 0/12/0/0
Oct 6 19:03:55 server1 vmunix: SCSI: Reset detected -- path: 0/12/0/1

I have checked the disks with ioscan, and in cstm if I map, I do not get any warnings.

What caused this, and how do I completely check this out?

As well, how do I clear out the I/O Errors in the cstm > logutil

Date/time of first entry: Wed Feb 9 19:01:40 2005

Date/time of last entry: Thu Oct 6 19:04:00 2005



Number of LPMC entries: 0
Number of System Overtemp entries: 0
Number of LVM entries: 0
Number of Logger Event entries: 0

Number of I/O Error entries: 519


Device paths for which entries exist:

(2) 0/12/0/0.3.0
(2) 0/12/0/0.2.0
(2) 0/12/0/0.1.0
(2) 0/12/0/0.0.0
(2) 0/10/0/0.2.0
(2) 0/10/0/0.0.0
(2) 0/10/0/0.3.0
(2) 0/10/0/0.1.0
(503) 0/0/2/0.0.0

Products for which entries exist:

(16) SCSI Disk
(503)

Product Qualifiers for which entries exist:

(2) HP18.2GMAP3367NC#HJ
(14) HP18.2GMAN3184MC
(503)

Logger Events for which entries exist:

(519) sdisk


I am going to assume that the (503) errors are from a previous issue we had that we did not clear up. I am hoping, since it makes no mention to that path in dmesg or syslog.
How do I reset?

[Steven E. Protter]

Once lbolt is detected it can only be reset by restarting the system.

If the lbolt was triggered by pulling and replacing a hotswap drive, then you can ignore it until the next scheduled boot. It will show up in dmesg until boot.

If there is still a bad disk or scsi problem it will have to be corrected to clear the warning.

SEP

[Ratzie]

There was nothing done to the system when this reset happened. How do I go about checking the drives

[Juan M Leon]

if you do
ioscan -funH 0/0/2/0.0.0
what do you get>

under cstm
map
sel dev (number)
update
il

what do you get.
what devices are attached to that hardware path.

Good luck

[Geoff Wild · Accepted solution]

Are you sure nothing physical happened?

This happened to one of my servers the other day - when a new server was placed into the same rack - and someone accidentally bumped one of the SCSI cables...

My error:

Oct 4 07:31:17 pc2000 vmunix: SCSI: Resetting SCSI -- lbolt: 2043443378, bus: 11
Oct 4 07:31:17 pc2000 vmunix: SCSI: Reset detected -- lbolt: 2043443378, bus: 11



To completely check the disks - do a dd - mine was:

dd if=/dev/dsk/c11t3d0 of=/dev/null


Rgds...Geoff

[Ratzie]

I have a print out from the cstm, looks ok except for non-medium errors

43 0/12/0/0.2.0 SCSI Disk (HP18.2GMAN3184 Information Successful
44 0/12/0/0.3.0 SCSI Disk (HP18.2GMAN3184 Information Successful
45 0/12/0/0.15.0 Disk Enclosure (HPA6491A)
46 0/12/0/1 PCI SCSI Interface (10000
47 0/12/0/1.8.0 SCSI Disk (HP18.2GMAN3184 Information Successful
48 0/12/0/1.9.0 SCSI Disk (HP18.2GMAN3184 Information Successful
49 0/12/0/1.10.0 SCSI Disk (HP18.2GMAN3184 Information Successful
50 0/12/0/1.11.0 SCSI Disk (HP18.2GMAN3184 Information Successful
51 0/12/0/1.15.0 Disk Enclosure (HPA6491A)
52 33 CPU (5df) Information Killed
53 37 CPU (5df) Information Successful
.... cerberus : 142.160.165.106 ....
55 101 CPU (5df) Information Successful
-- Information Tool Log for PCI SCSI Interface on path 0/10/0/0 --Successful
cstm>update
Log creation time: Fri Jul 30 21:35:13 2004
-- Converting multiple raw log files to text. --
Hardware path: 0/10/0/0on Tool Log for each selected device...


Product ID: PCI SCSI Interface
Device ID: 0x0021
Revision ID 0x0001
Vendor ID: 0x1000 ( Symbios Logic Inc.)
Class Code: 0x010000
Base Class: 0x01 ( Mass Storage Controller. )
Sub-Class/Interface: 00/00 ( SCSI bus controller )
Device Status: 0x0230
Bit 5: 66 MHz Capable
Bit 9-10: DEVSEL timing 01 - medium

=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=

-- Information Tool Log for SCSI Disk on path 0/10/0/0.0.0 --

Log creation time: Fri Sep 9 05:00:34 2005

Hardware path: 0/10/0/0.0.0

Error Logs
Total Retries: 0 Buffer Overruns: N/A
Read Reverse Errors: N/A Buffer Underruns: N/A
Write Errors: 0 Non-Medium Errors: 50
Verify Errors: 0
Capacity (M Byte): 17366.45
=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=-+-=
Max Block Address: 35566479
-- Information Tool Log for SCSI Disk on path 0/10/0/0.1.0 --

Log creation time: Fri Sep 9 05:00:34 2005

Hardware path: 0/10/0/0.1.0


Product Id: MAN3184MC Vendor: HP 18.2G
Device Type: SCSI Disk Firmware Rev: HP05
Device Qualifier: HP18.2GMAN3184MC Logical Unit: 0
Serial Number: MN136072
Capacity (M Byte): 17366.45
Block Size: 512
Max Block Address: 35566479
Error Logs
Total Retries: 0 Buffer Overruns: N/A
Read Reverse Errors: N/A Buffer Underruns: N/A
Write Errors: 0 Non-Medium Errors: 2277
Verify Errors: 0


Product Id: MAN3184MC Vendor: HP 18.2G
Device Type: SCSI Disk Firmware Rev: HP05 --
Device Qualifier: HP18.2GMAN3184MC Logical Unit: 0
Serial Number: MN136045 0:35 2005
Capacity (M Byte): 17366.45
Block Size: 512
Max Block Address: 35566479
Error Logs
Total Retries: 0 Buffer Overruns: N/A
Read Reverse Errors: N/A Buffer Underruns: N/A
Write Errors: 0 Non-Medium Errors: 49
Verify Errors: 0

The ioscan returned everything claimed.

I did notice that in map, I have an information killed, and when I try to update and il
I get infolog is currently disabled.
I should have 4 good cpu's

[Ratzie]

Thanks for the help