Data recovery problem on Smart Array 5300

[Florin Bota]

Hello all,
I have a problemm with a Proliant ML350 G3 server with a Smart Array
5300 SCSI RAID controller. There are 6 SCSI disks connected to the
controller, 5 of them configured in RAID 5 with one spare disk. At some
moment one of the disks failed, so I replaced it. After that each time
the server is starting I get a message indicating that Automatic Data
Recovery is needed, choose F1 (do ADR) and the server works fine.
However on the new disk the status light remains off (so no steady
green light). I investigated and it seems ADR can not finalize
corectlly due to a data read error on another disk. This is a situation
indicated in the documentation
(http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?lang=en...
- SMART5300- POST Messages) and there they indicate to do a full backup
of readable data and run "Diagnostics Surface Analysis" and restore
afterwards. Where can i find that programm? Is is part of some HP
Diagnostics, Array Configuration Utility or what? If anyone used that
Diagnostics Surface Analysis can please indicate me in which programm I
can find this one, or how to run it (maybe I have to use the SmartStart
CD? boot in some special mode?).
Is that Surface Analysisi alike a scandisk looking for bad sectors
on the surface and recovering them, or I will have to completlly
recover the array (create a new one and fully reinstall the server,
restore data after)?

Thanks a lot,
Florin.

[Oleg Koroz]

http://h18023.www1.hp.com/support/files/server/us/index.html
Find your system O/S Generation
Utility that you need ADU
Array Diagnostic Utility

upload txt file back to ITRC

[Florin Bota]

Hello, here is the report generated by the Array Diagnostics Utility.
Thank you,
Florin.

[Oleg Koroz]

Firmware Updates Needed
BD072863B2 HPB5 newer HPB8 A
BD07288277 HPB0 ----- HPB4 A
BD072863B2 HPB5 ----- HPB8 A
BD0728856A HPB0 ----- HPB2 B SCSI Port 2, Drive ID 3 has Read write Problems
BD072863B2 HPB5 ----- HPB8 A
BD07287B4C HPB7 ----- HPBC A


Scan for Bad sectors it appears ID 3 need replacement
Rebuild failed due Read Error

Also Update Firmware for all HDD
http://h18023.www1.hp.com/support/files/server/us/download/24777.html

If bad sectors found replace ID3 and reinstall all from scratch, new O/S

a few errors from SCSI chan that might be related to controller or SCSI backplane, however after Firmware Update this might get changed

From ACU you have any entry for Unused or Unlocated space ?

[Florin Bota]

No free space on that array, all assigned to one logical disk.
How should i run the bad sector scan? Is there a specific programm I have to use?
To update the firmware is it ok to use the last Firmware Maintenante CD available on hp site for the ML350 G3?
Doing a ghost type image of the logical disk, recreating the array and applying the previous created image to the new array can solve the problemm?
Thank you very much for the help,
Florin.

[Oleg Koroz]

if you can make a good image where no corrupted files, .. go for it.

http://www.ss64.com/nt/chkdsk.html

[Florin Bota]

I thought I have to use some special HP program instead of the microsoft's one (which sees only the logical disk and not the physical HDDs inside the server). I'll doo a full backup, update firmware, do a ghost-type image and try to run chkdsk after. If it will not solve the problemm and bring the array back on-line, I will destroy the array, recreate it and apply the ghost image back ... Hope it will finally solve the problemm and have the array back on-line (since now I think that one disk broking would lead to all data loss, since the array state is "Ready to rebuild"...
Thanks again,
Florin.