cstm infolog issue

[yyghp]

I can see the following data from the "cstm infolog" for the hard disk:

Hardware path: 10/8.3.0
Product Id: ST34572WC Vendor: SEAGATE
Device Type: SCSI Disk Firmware Rev: HP03
Device Qualifier: SEAGATEST34572WC Logical Unit: 0
Serial Number: JK9836300GEB2K
Capacity (M Byte): 4095.86
Block Size: 512
Max Block Address: 8388313
Error Logs
Read Errors: 7 Buffer Overruns: N/A
Read Reverse Errors: N/A Buffer Underruns: N/A
Write Errors: 116 Non-Medium Errors: 2

There are some read and write errors. Is that something wrong with the disk or is it normal?

There is no events in syslog.log

Thanks!

[James R. Ferguson]

Hi:

This may be stale information. You can test the disk using this:

# dd if=/dev/rdsk/cXtYdz of=/dev/null bs=256k

Note the use of a raw device such that the LVM layer (and buffer cache) is bypassed. The above will read the entire disk. Look for reported errors at the end and/or any EMS alerts or syslogd information.

Regards!

...JRF...

[Michael Steele_2]

Go to logtool and note the time interval between errors. If the time interval is short, like 24 hours, call in HW. But is its extended over several weeks, forget about it.

STM > TOOLS > UTILITY > RUN > LOGTOOL > FILE > VIEW > RAW SUMMARY

[Albert_31]

Hi,

The disk you are using is a very OLD one..4GB which does not have teh in built capabilities of detecting and reporting h/w faults, So I guess you should check the disk state using the dd command specified earlier/log a hw call.

also check if the errors are increasing

# echo "map selall info; wait infolog" | cstm > output file.. check if the errors are increasing as well..to take the decision on disk replacement or not.

regards

[Andrew Merritt_2]

If you've got the OnlineDiags installed, and 'disk_em' is running, you should get an EMS event generated if the level of errors reaches a level that should cause concern.

Some errors are expected on disks these days, and a low level is acceptable as long as the errors are recoverable. The disk monitor has this knowledge built in, for older disks.

Newer disks (9Gb and bigger) use SMART technology, which will cause EMS events (via disk_em) itself when error rates reach a certain level.

I would recommend that you check that you have an up-to-date version of the OnlineDiags.

Andrew

[Scott Donaldson]

Hi,

Take action soon or at least check your service contract - if you have one that covers that drive. My hot pluggable drives (A3647A ?) in a D380 use the same Seagate drive mechanism and 'fall off' of my support contract on June 30, 2006 (in Canada).

Just to provide you with some reassurance one of those drives just up and died last week without logging any read/write errors at all. Of course the drive failed 30 seconds after midnight on 6606?!!

Scott.