New pSeries Server with Disk Error (errpt)

[KOG]

Hi Folks

Need second opinion here, the new server arrived yesterday and I ran errpt this morning and found there is a problem with the hdisk1 (it has P state which concerns me)-

# errpt
IDENTIFIER TIMESTAMP T C RESOURCE_NAME DESCRIPTION
9DBCFDEE 0402092104 T O errdemon ERROR LOGGING TURNED ON
192AC071 0402092104 T O errdemon ERROR LOGGING TURNED OFF
16F35C72 0402064304 P H hdisk1 DISK OPERATION ERROR
16F35C72 0401224304 P H hdisk1 DISK OPERATION ERROR
C291A40C 0401210004 P O dumpcheck Dump compression was turned on.
A6DF45AA 0401145004 I O RMCdaemon The daemon is started.
2BFA76F6 0330163404 T S SYSPROC SYSTEM SHUTDOWN BY USER
9DBCFDEE 0401134404 T O errdemon ERROR LOGGING TURNED ON
192AC071 0330163304 T O errdemon ERROR LOGGING TURNED OFF
A6DF45AA 0330155404 I O RMCdaemon The daemon is started.
2BFA76F6 0329143504 T S SYSPROC SYSTEM SHUTDOWN BY USER
9DBCFDEE 0330155304 T O errdemon ERROR LOGGING TURNED ON
192AC071 0329143404 T O errdemon ERROR LOGGING TURNED OFF

Should I make a complaint to the company where we purchased the server? Or is there any further tests I should run?

Also I have noticed a new feature on AIX 5L with regarding to /opt filesystem, it is usually used for oracle system files and should be used within oracle vg so ques I am asking is if it is an 'optional' filesystem within Aix 5L is it ok for me to remove it and create /opt within oracle vg so that I can install oracle system files?

Thanking you all in advance.

Regards

K

 

[unixfreak]

Please send the output of 'errpt -aj 16F35C72'

Cheers

 

[KOG]

Hi

Many thanks for your quick response, here's the output (excuse for the length of it)

---------------------------------------------------------------------------
LABEL: DISK_ERR2
IDENTIFIER: 16F35C72

Date/Time: Fri 2 Apr 06:43:59 2004
Sequence Number: 26
Machine Id: 0059D05F4C00
Node Id: blb6e3
Class: H
Type: PERM
Resource Name: hdisk1
Resource Class: disk
Resource Type: scsd
Location: U0.1-P1/Z1-A3
VPD:
Manufacturer................IBM
Machine Type and Model......ST336605LC
FRU Number..................09P4445
ROS Level and ID............43353042
Serial Number...............00025A1A
EC Level....................H11936
Part Number.................09P4444
Device Specific.(Z0)........000003129F00013E
Device Specific.(Z1)........0501C50B
Device Specific.(Z2)........1000
Device Specific.(Z3)........02121
Device Specific.(Z4)........0001
Device Specific.(Z5)........22
Device Specific.(Z6)........162870 C

Description
DISK OPERATION ERROR

Probable Causes
DASD DEVICE

Failure Causes
DISK DRIVE
DISK DRIVE ELECTRONICS

Recommended Actions
PERFORM PROBLEM DETERMINATION PROCEDURES

Detail Data
PATH ID
0
SENSE DATA
0A03 0000 2A00 01D2 32A8 0000 0800 0000 0102 0000 7000 0100 0000 000A 0000 0000
5D00 4300 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0014 0008

Diagnostic Analysis
Diagnostic Log sequence number: 16
Resource tested: hdisk1
Resource Description: 16 Bit LVD SCSI Disk Drive
Location: U0.1-P1/Z1-A3
SRN: 59B-130
Description: Error log analysis indicates problem reported by disk
drive's self monitoring function.
Possible FRUs:
hdisk1 FRU: 09P4445 U0.1-P1/Z1-A3
16 Bit LVD SCSI Disk Drive

---------------------------------------------------------------------------
LABEL: DISK_ERR2
IDENTIFIER: 16F35C72

Date/Time: Thu 1 Apr 22:43:22 2004
Sequence Number: 25
Machine Id: 0059D05F4C00
Node Id: blb6e3
Class: H
Type: PERM
Resource Name: hdisk1
Resource Class: disk
Resource Type: scsd
Location: U0.1-P1/Z1-A3
VPD:
Manufacturer................IBM
Machine Type and Model......ST336605LC
FRU Number..................09P4445
ROS Level and ID............43353042
Serial Number...............00025A1A
EC Level....................H11936
Part Number.................09P4444
Device Specific.(Z0)........000003129F00013E
Device Specific.(Z1)........0501C50B
Device Specific.(Z2)........1000
Device Specific.(Z3)........02121
Device Specific.(Z4)........0001
Device Specific.(Z5)........22
Device Specific.(Z6)........162870 C

Description
DISK OPERATION ERROR

Probable Causes
DASD DEVICE

Failure Causes
DISK DRIVE
DISK DRIVE ELECTRONICS

Recommended Actions
PERFORM PROBLEM DETERMINATION PROCEDURES

Detail Data
PATH ID
0
SENSE DATA
0A03 0000 2A00 01D2 3508 0000 0800 0000 0102 0000 7000 0100 0000 000A 0000 0000
5D00 4300 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0013 0004

Diagnostic Analysis
Diagnostic Log sequence number: 10
Resource tested: hdisk1
Resource Description: 16 Bit LVD SCSI Disk Drive
Location: U0.1-P1/Z1-A3
SRN: 59B-130
Description: Error log analysis indicates problem reported by disk
drive's self monitoring function.
Possible FRUs:
hdisk1 FRU: 09P4445 U0.1-P1/Z1-A3
16 Bit LVD SCSI Disk Drive


Many thanks

Regards

K

 

[unixfreak]

This means it has multiple unrecoverable errors. Call IBM to have it replaced.
You can give them the part of the sense data that says "5D00 4300" and see what they say.

Cheers

 

[uptodate]

Yes, call IBM or the VAR and have the disk replaced. It is possible that the FRU for this disk has been recalled due to engineering problems and is replaced by a new disk.

 

[MoshiachNow]

Hi,

Actually ,error 5D00 means "5D/00h FAILURE PREDICTION THRESHOLD EXCEEDED".

Possibly (not nessesarly) you could get away with a physical reformating of the disk using the diagnostics CD.

"Long live king Moshiach !"

http://www.7for70.com/

 

[KOG]

Many thanks to all of you for your responses, I am expecting a consultant from the company where the compay bought the server .. he is going to check and run h/w diag so will leave the decision to him.

Thanks again

Regards

K

 

[bjverzal]

As for '/opt/' - I'd leave it alone. You could try moving it to another VG, but next 'migration' could cause problems.

BV

 

[bi]

regarding /opt in the Oracle Volume Group: You could create an LV with a filesystem named /opt/oracle (or /opt/whatever) and it can be in the Oracle volume group. You can then install Oracle into /opt/oracle. The software won't know the difference, but you will still have your Oracle outside of rootvg.

But I agree with BV: best to leave default filesystems (like plain old /opt) alone.