Extended Sense Data from Device

[CCH_3]

We have few Itanium RX2620 server with OpenVMS 8.2-1 always encounter Extended Sense Data from Device. How to solve it? Pls refer to the attachment for full error log details. Thanks.

[Steven Schweda]

> [...] refer to the attachment [...]

Does OpenVMS 8.2-1 create a Microsoft Word
document for its error messages, or was that
your idea? If the data started out as plain
text, then why not leave them as plain text?
Everyone can read plain text. Not everyone
can read a Microsoft Word document.

[Richard Brodie_1]

The most useful information is often the ASC/ASCQ:

x47, x3 -> (SCSI) Bus CRC error.

Cabling/seating would seem like the obvious place to start looking.

HP also tend to provide fairly frequent updates to controller and drive firmware (amongst other things). I don't suggest you go out and update now; it might be worth reviewing your current revisions though.

[CCH_3]

Event: VMS Device Error Event occurred at Sun 5 Dec 2010 18:12:07 GMT+10:00
COMMON EVENT HEADER (CEH) V2.0 Event_Leader xFFFF FFFE
Header_Length 284
Event_Length 704
Header_Rev_Major 2
Header_Rev_Minor 1
OS_Type 2 -- OpenVMS
Hardware_Arch 6 -- Itanium®-based computers
CEH_Vendor_ID 3,564 -- Hewlett-Packard Company
Hdwr_Sys_Type 64 -- Intel® Itanium® architecture
Logging_CPU 0 -- CPU Logging this Event
CPUs_In_Active_Set 1
Major_Class 1
Minor_Class 1
Entry_Type 1,001 -- VMS Device Error Event
DSR_Msg_Num 0 -- Unused
Chip_Type 31 -- Itanium® 2 processor
CEH_Device 54
CEH_Device_ID_0 x0000 0000
CEH_Device_ID_1 x0000 0000
CEH_Device_ID_2 x0000 0000
Unique_ID_Count 6,098
Unique_ID_Prefix 1,510
Exact_Length 409
Num_Strings 6

TLV Section of CEH TLV_DSR_String HP rx2620 (1.30GHz/3.0MB
TLV_DDR_String HP 73.4G ST373453LC
TLV_Sys_Serial_Num SGH46150V1
TLV_Time_as_Local Sun 5 Dec 2010 18:12:07 GMT+10:00
TLV_OS_Version V8.2-1
TLV_Computer_Name SYSC

Entry_Type 1,001
EMB_Block emb_ertcnt 16 Error Count
emb_ertmax 16 Max error count
emb_iosb 0
emb_sts x1802 1810
emb_class 1 Disk Class
emb_type 54
emb_rqpid 65,642
emb_boff 2,048
emb_bcnt 63,488
emb_media 7,195,652
emb_unit 200
emb_errcnt 8
emb_opcnt 80,474
emb_ownuic x0001 0004
emb_char x1C4D 4008
emb_Device_Number 0
emb_func 8,203
emb_name_len 8
emb_name SYSC$DKB
emb_dtname_len 19
emb_dtname HP 73.4G ST373453LC
Generic DK Driver Header Revision 2 HW_Revision HPC5
DK_Error_Type x05 Extended Sense Data from Device
DK2_SCSI_ID 2
DK2_SCSI_LUN 0
DK2_SCSI_SubLUN 0
DK_Port_Status x0000 0001 Normal Successful Completion
DK2_SCSI_CMD_Len 10
DK2_SCSI_CMD Dump starting at offset: x1a8
[x0] x6d00002a
[x4] x004cc
[x8] x000
SCSI_Status 2 Check Condition
DK2_Additional_Data_Len 18
DK_Sense_Data Sense_byte_0 xF0
Error_Code[6:0] x70 Current Error
Valid[7] x1 Valid
Segment_Num 0
Sense_Key x0B
Sense_Key[3:0] xB Abort Command
Info_bytes x006D CC04 Usually LBA
Additional_Sense_Len 10 Extended Sense Data Length
Cmd_spec_info x0000 0000
ASCQ_ASC x0347
ASC[7:0] x47
ASCQ[15:8] x3
FRU_code x08
Sense_Key_Specfic_Byte0 x00
Bit_Pointer[2:0] x0
BPV[3] x0 BPV Invalid
C/D[6] x0 Data parameters
Sense_Key_Specific_Valid[7] x0 Invalid
Sense_Key_Specfic_Bytes 0

[Hoff]

Generic extended sense data reception is normal.

As for the x47/x00 code, that's a SCSI parity error.

Possibly entirely transient.

A bus parity error can also indicate a hardware or cabling or termination error, loose connection, signal transient, power supply problem, etc. This particularly if these or related errors are repeating or particularly increasingly frequent, or are involving other devices on this SCSI bus.

The obvious inferences in aggregate imply that this error might be reoccurring.

Which leads to the generic answer for all hardware errors...

Contact your hardware support provider for assistance.

Confirm that you have consistent and full backups of this disk data, too. (I wouldn't expect to need it here, but it's useful to have a good backup any time somebody goes near the hardware.)

Also consider whether RAID-1 is appropriate; host-based volume shadowing (HBVS) or bus-based or storage controller-based RAID mirroring.

And for completeness, your operating system software is quite stale. This could well be a software bug; there have been a number of updates to the SCSI driver stack since your release.

[Hoff]

Correction, mis-read on my part, that's 47/03 and not 47/00, and that 47/03 is...

"INFORMATION UNIT iuCRC ERROR DETECTED"

That's a CRC detected between the host and the widget.

Same triggers as before, however. Check the SCSI cable, bus length, improper termination, HBA problems, another device, etc., and call your preferred hardware service vendor.

Reposting due to ITRC glitch.