View previous topic :: View next topic |
Author |
Message |
treffer Apprentice
Joined: 14 Dec 2004 Posts: 150
|
Posted: Wed Aug 09, 2006 10:29 am Post subject: [SOLVED] SCSI Hitachi Ultrastar // 516bytes sector size |
|
|
Hi,
I have four 36GB Hitachi Ultrastar DK32EJ-36NW with a Symbios controller. (U2W/LVD)
The drives are correctly dectected and doing the spinup, direkt disk format works fine, but it's impossible to use the drives because of the 516 bytes sector size (see dmesg output [1]).
scsiinfo [2] shows settings that differ from the Manufactory defaults [3]. However the drive refuses to change the sector size (scsiinfo -fXR does not give an error but does not change anything, we've also tried sg_info and the controller format util which should change sector size). The drives were part of a raid system, it might be the same problem as described here: help needed with fibre channel disk (how can I format?)
Can anyone tell me
- How to set the sector size with a software tool or
- How to clear the nvram/reset the drive parameters or
- where to get a copy of the working firmware for the device to flash the drive (and how to flash it)
There is no important data on the drive (several direct drive formats passed).
Thanks in advance.
[SOLUTION]
We have contacted the Hitachi Support and they gave us everything we needed!
[1]
dmesg
Code: |
sym0: <895a> rev 0x1 at pci 0000:00:09.0 irq 6
sym0: Symbios NVRAM, ID 15, Fast-40, LVD, parity checking
sym0: open drain IRQ line driver, using on-chip SRAM
sym0: using LOAD/STORE-based firmware.
sym0: handling phase mismatch from SCRIPTS.
sym0: SCSI BUS has been reset.
scsi0 : sym-2.2.3
target0:0:1: Multiple LUNs disabled in NVRAM
Vendor: HITACHI Model: DK32EJ-36NW Rev: K1K1
Type: Direct-Access ANSI SCSI revision: 03
target0:0:1: tagged command queuing enabled, command queue depth 16.
target0:0:1: Beginning Domain Validation
target0:0:1: asynchronous
target0:0:1: wide asynchronous
target0:0:1: FAST-40 WIDE SCSI 80.0 MB/s ST (25 ns, offset 31)
target0:0:1: Domain Validation skipping write tests
target0:0:1: Ending Domain Validation
sda : unsupported sector size 516.
SCSI device sda: 0 512-byte hdwr sectors (0 MB)
sda: Write Protect is off
sda: Mode Sense: c3 00 10 08
SCSI device sda: drive cache: write through w/ FUA
sd 0:0:1:0: Attached scsi disk sda
target0:0:4: Multiple LUNs disabled in NVRAM
target0:0:8: Multiple LUNs disabled in NVRAM
|
[2]
scsiinfo -f /dev/sda
Code: |
Data from Format Device Page
----------------------------
Removable Medium 0
Supports Hard Sectoring 1
Supports Soft Sectoring 0
Addresses assigned by surface 0
Tracks per Zone 7665
Alternate sectors per zone 14910
Alternate tracks per zone 21
Alternate tracks per lun 336
Sectors per track 710
Bytes per sector 516
Interleave 1
Track skew factor 83
Cylinder skew factor 155
|
[3]
scsiinfo -fM /dev/sda
(-M = Manufactory defaults)
Code: |
Data from Format Device Page
----------------------------
Removable Medium 0
Supports Hard Sectoring 1
Supports Soft Sectoring 0
Addresses assigned by surface 0
Tracks per Zone 7665
Alternate sectors per zone 14994
Alternate tracks per zone 21
Alternate tracks per lun 336
Sectors per track 714
Bytes per sector 512
Interleave 1
Track skew factor 84
Cylinder skew factor 156
|
[EDIT #1]
Substituted missleading "lowlevel format", the term was used in a wrong way.
[EDIT #2]
Changed subject to includes [solved]
[EDIT #3]
added [SOLUTION] _________________ root@localhost# whois POEM-RIPE55-SONG
root@localhost# : ( ) { : | : & } ; :
Last edited by treffer on Tue Aug 15, 2006 4:32 pm; edited 3 times in total |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54424 Location: 56N 3W
|
Posted: Wed Aug 09, 2006 10:52 am Post subject: |
|
|
treffer,
You cannot actually do a low level format (write sector and track address marks) to a voice-coil head actuator drive. All drives have been made that way for a long time now. Low level format information is held on a read only platter, along with the track alignment information that allows the heads to be dynamically repositioned on the track as trhe drive rotates.
All a low level format does now is reads the disk.
When voice-coil head actuator drives were first introduced to PCs, the servo platter (on some drives) was not read only, this was unfortunate for the users that attempted a low level format, since it erased the servo information, rendering the drive useless.
Its unlikely a drive firmware change will fix this for you, although, I suppose it could be made to lie about the sector size and cover up for the actual larger sectors. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
treffer Apprentice
Joined: 14 Dec 2004 Posts: 150
|
Posted: Wed Aug 09, 2006 1:50 pm Post subject: |
|
|
Thanks for your reply.
I didn't believe it is trivial, however it should be possible according to the technical specification (http://www.hitachigst.com/tech/techlib.nsf/techdocs/48BDF449EE74698A86256EE70077520C, DK32EJ-NC/NW Product Specifications):
Quote: |
...
Bytes/sector (see Note 3) 512, 516, 520, 524, 528,1024 Bytes
...
Note 3. DK32EJ is formatted to 512 Bytes/sector as the default.
If non-512 Bytes/sector is required, customer can change it by re-format
of the drive.
|
So several different byte-per-sector layouts should be supported and it should be possible to change the setting. However changing the sector size fails.
By lowlevel format I meant the direct hd format done by the controler or the hitachi tools without respecting any partition table or the mbr.
The term was used for a very special format on older disks, I wasn't aware of that, so my statement on that was simply wrong! Thanks for the correction.
Any further ideas/hints? _________________ root@localhost# whois POEM-RIPE55-SONG
root@localhost# : ( ) { : | : & } ; : |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54424 Location: 56N 3W
|
Posted: Wed Aug 09, 2006 4:30 pm Post subject: |
|
|
treffer,
Being a hardware guy, my next step would be to connect a bus analyser to the scsi bus and see if it was actially getting the format device commad that you are sending, or if the sector size is being ignored.
The next best thing, is to try another plain ordinary SCSI adaptor, in case the format device SCSI command is being intercepted somehow. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
treffer Apprentice
Joined: 14 Dec 2004 Posts: 150
|
Posted: Thu Aug 10, 2006 9:44 am Post subject: |
|
|
Ok, but the controller actually prefers a 512bytes/sector and tries to change the sector size as soon as you try to format the drive with the controller. Additionally reading drive parameters, doing a drive check or formating the drive (with the hitachi tools) works flawless (and the drive's led indicates drive activity).
So I expect the commands to reach the drive. Otherwise I can't understand why most of the basic stuff works (except deleting the mbr - which schould be 512 bytes anyway). Especially reading the scsi modes works, but not writing. Is there a chance that a hardware related problem could block these commands (the read-only jumper of the drive is not set)? In this case I would try to format the drive with another known-good controller/cable pair...
Thanks in advance. _________________ root@localhost# whois POEM-RIPE55-SONG
root@localhost# : ( ) { : | : & } ; : |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54424 Location: 56N 3W
|
Posted: Thu Aug 10, 2006 10:00 am Post subject: |
|
|
treffer,
After you use the SCSI format command, do you need to power cycle the drive to get it to report the new block size? _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
treffer Apprentice
Joined: 14 Dec 2004 Posts: 150
|
Posted: Tue Aug 15, 2006 4:19 pm Post subject: Solved |
|
|
We have finally contacted the Hitachi Support and they gave us a special tool to change the sector size. The whole procedure isn't trivial, and the tool is rather complicated (aka not for end users), but it worked after some try and error
So if you run into the same problem with your hitachi drive, contact the hitachi support, they will help you - seriously! I was stunned by the quality of there support - we got a working solution within two working days - wow, great job!
We have still hundreds of Seagte Cheetah 9GB drives with the same problem... [yes, literally hundreds]
Thanks for all replies! _________________ root@localhost# whois POEM-RIPE55-SONG
root@localhost# : ( ) { : | : & } ; : |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|