Install Clean HDD in (white light) My Book World Edition / Discussion

This is the discussion related to the wiki page Install Clean HDD in (white light) My Book World Edition.

page 1 of 212next »

Success!

gijsh 29 Dec 2009 12:07

I unbricked my Mybook WE White Light. Thank you for you detailed help. Other pages on this site were not as helpful; the disk booted, but gave always an error message about DataVolume not found. So no shares, and no useable disk space.

FYI, I followed your instructions literally, with a few additions/exceptions:

• I used a newer version of the firmware, downloaded from http://websupport.wdc.com/firmware/list.asp?type=WDH1NC&fw=01.00.02. The decoding as described here went errorless.
• I explicitly formatted /dev/sdb4 as XFS, the latter being the preferred file system for WD.

And now for some better software (optware or other) to make the thing still more useful.

Re: Success!

tomas123 02 Jan 2010 20:25

it's nice to hear, that the guide work

> I explicitly formatted /dev/sdb4 as XFS

in the guide i use the simplest way, to make XFS and all the shares

touch ./sdb1/etc/.factory_restore

this flag runs at next boot the script /tools/factoryDefault.sh (see script /etc/init.d/rcS)
a part of the script factoryDefault.sh is

echo -n "=====>Making xfs filesystem....."

But if you have a broken disk you can also restore the firmware and save your old datas on sda4 if you delete the flag ./sdb1/etc/.factory_restore

No luck :(

yvlasenko 28 Mar 2010 15:44

Hi, Tomas!

After your comment I've tryed everything again (erasing the whole drive with "dd if=/dev/zero of=/dev/sdc"), but no luck.

The only error was that partition were not mounted:

mdadm: error opening /dev/md_d0: No such file or directory
mdadm: error opening /dev/md_d1: No such file or directory
mdadm: error opening /dev/md_d3: No such file or directory
mdadm: Unrecognised md component device - /dev/sdc1
mdadm: Unrecognised md component device - /dev/sdc2
mdadm: Unrecognised md component device - /dev/sdc3

And there were warnings of mdadm:

mdadm: Note: this array has metadata at the start and
    may not be suitable as a boot device.  If you plan to
    store '/' or '/boot' on this device please ensure that
    your boot-loader understands md/v1.x metadata, or use
    --metadata=1.0
Continue creating array? mdadm: Defaulting to version 1.1 metadata
mdadm: array /dev/md1 started.

I'm not sure if this matters.

I've tryed the same with "—metadata=1.0" option, no luck again.

Re: No luck :(

tomas123 28 Mar 2010 17:20

The solution it's simple. Every linux do one's own thing by naming the raid arrays.
My walktrough is for Ubuntu 9.10, as I wrote.

Your RAID name is /dev/md? and not /dev/md_d?

Correct the commands
Awaiting to hear from you

Last edited on 28 Mar 2010 17:21 by tomas123 Show more

Re: No luck :(

yvlasenko 28 Mar 2010 18:04

It is mdX in my case. But the only place the wrong name appears is "mdadm —stop /dev/md_d0", when there is no raid devices (my Arch has automount disabled). In other places we call mdX, as in the guide ("echo "y" | mdadm —create /dev/md0 -l 1 -n 2 /dev/sdc1 missing", for example).

Re: No luck :(

tomas123 28 Mar 2010 19:06

you are right, I'm answered too fast

I checked the script 10 times with two different drives.
User Gijsh also announce succes above.
I think, there is a minor different in your linux . What's your linux?
I'm using ubuntu 9.10 in vmware with a physical disc (for advanced user)

Here is my original protocol with all messages:
http://pastebin.com/4dVv8s8x

maybe you find the different.

A hint for debug: you can first discard the command

touch ./sdb1/etc/.factory_restore

Then the whitebook starts with a nativ Web-GUI (error message about missing sda4)

some trip hazards:
Your bash process this command correct? echo -ne '\x00\x00\x30' | dd of=/dev/sdb seek=255
Do you use the original disc?
Do you have some trouble with the MBR (non original disc)?

Re: No luck :(

yvlasenko 29 Mar 2010 14:45

Well, after playing for some time I'll have to give up. Looks to me I faced h/w problem.

Tried WDRescueTool too, the same result.

Thanks anyway.

Re: No luck :(

tomas123 29 Mar 2010 16:50

thank you for feedback

Help Please

samcargo 06 Jul 2010 06:40

Having followed carrefully all the commands i succeeded thanks to you install a new samsung 1T hdd on my mbwe
however upon accessing the console i get this:

07/06 14:18:51 [Volume Status] Data Volume 'DataVolume' doesn't exist!

I can login but can't see not data volume, here is what my disk looks like:

Home | Volume Status
Volume Type Disk Usage Size Status
DataVolume JBOD unknown unknown Failed

Please help me.

Re: Help Please

tomas123 06 Jul 2010 11:14

You have a samsung disk.
Have you followed the part for users of non WD disk?
You must hack the factoryDefault.sh…

Re: Help Please

samcargo 06 Jul 2010 17:46

Thank you for your reply,

I did it as mentioned

gedit ./sdb1/proto/SxM_webui/admin/tools/factoryDefault.sh

cat -n ./sdb1/proto/SxM_webui/admin/tools/factoryDefault.sh | head -161l | tail -5l

1. 157 cat /sys/block/sda/device/model | grep WDC >/dev/null 2>&1
2. 158 #if [ "$?" -eq "0" ]; then
3. 159 let WD_DISK_COUNT=WD_DISK_COUNT+1
4. 160 DISK1=sda
5. 161 #fi

However the disk partition sda4 which has the biggest size left more or less 900 GB is not recognized as data volume.
Mind you my MBWE board is originally at 2TB board and the disk i am using is a 1TB Samsung.

Re: Help Please

tomas123 06 Jul 2010 21:03

ok, I need more infos. I have a 1TB disk.

# dmesg
# sfdisk -l
# df
# cat /proc/partitions
# cat /proc/mounts

…but first make a factory reset over the web interface.
The factory reset formats the big partition.
If this failed then check your script /proto/SxM_webui/admin/tools/factoryDefault.sh if there a part for the 2TB Western Digital disk…

Last edited on 06 Jul 2010 21:07 by tomas123 Show more

Re: Help Please

samcargo 07 Jul 2010 00:42

Thank your for your help,
factoryDefault.sh

Is different than yours:

if [ "$MODEL" != "WWLXN" ]; then
let DISK_COUNT=DISK_COUNT+1
cat /sys/block/sdb/device/model | grep WDC >/dev/null 2>&1

1. if [ "$?" -eq "0" ]; then

let WD_DISK_COUNT=WD_DISK_COUNT+1
DISK2=sdb

1. fi

So i modified from sdb to sda

if [ "$MODEL" != "WWLXN" ]; then
let DISK_COUNT=DISK_COUNT+1
cat /sys/block/sda/device/model | grep WDC >/dev/null 2>&1

1. if [ "$?" -eq "0" ]; then

let WD_DISK_COUNT=WD_DISK_COUNT+1
DISK2=sda

1. fi

root@ubuntu:/home/ubuntu# dmesg

root@ubuntu:/home/ubuntu# sfdisk -l

Disk /dev/sda: 121601 cylinders, 255 heads, 63 sectors/track
Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from 0

Device Boot Start End #cyls #blocks Id System
/dev/sda1 4 247 244 1959930 fd Linux raid autodetect
/dev/sda2 248 279 32 257040 fd Linux raid autodetect
/dev/sda3 280 402 123 987997+ fd Linux raid autodetect
/dev/sda4 403 121600 121198 973522935 fd Linux raid autodetect

root@ubuntu:/home/ubuntu# df
df: cannot read table of mounted file systems: Input/output error

root@ubuntu:/home/ubuntu# cat /proc/partitions_

major minor #blocks name

7 0 687976 loop0
7 1 4188160 loop1
8 0 976762584 sda
8 2 257040 sda2
8 16 7818152 sdb
8 17 7818016 sdb1
root@ubuntu:/home/ubuntu#

root@ubuntu:/home/ubuntu# # cat /proc/mount
root@ubuntu:/home/ubuntu#

Nothing happens with the last command

Thank you again very very much
sam cargo

Last edited on 07 Jul 2010 07:47 by samcargo Show more

Recent Update commands used on ubuntu 9.10

samcargo 07 Jul 2010 03:15

root@ubuntu:/home/ubuntu# cat /proc/partitions
major minor #blocks name

7 0 684016 loop0
7 1 4188160 loop1
8 0 976762584 sda
8 1 1959930 sda1
8 2 257040 sda2
8 3 987997 sda3
8 4 973522935 sda4
8 16 7818152 sdb
8 17 7818016 sdb1
root@ubuntu:/home/ubuntu#

root@ubuntu:/home/ubuntu# cat /proc/mount
cat: /proc/mount: No such file or directory
root@ubuntu:/home/ubuntu#

Last edited on 07 Jul 2010 07:46 by samcargo Show more

Re: Recent Update commands used on ubuntu 9.10

samcargo 07 Jul 2010 03:21

Volume Type Disk Usage Size Status
DataVolume Span unknown unknown Failed

Re: Recent Update commands used on ubuntu 9.10

tomas123 07 Jul 2010 07:44

please remove the long dmesg output from your ubuntu PC ;-)
you can upload it to http://ifile.it/

I think your problem is a wrong patch in factoryDefault.sh for your disk SAMSUNG HD103UJ

This is a 1TB disk - you wrote 2TB?

Sorry, but I haven't enough time to solve your problem.
Read your factoryDefault.sh carefully and find the part to beat the WD/Samsung disk barrier.
In your single disk WhiteBook the SAMSUNG HD103UJ is sda !!
(sdb is the second disk in the 2 disk whitebooks)

Last edited on 07 Jul 2010 07:54 by tomas123 Show more

Re: Recent Update commands used on ubuntu 9.10

tomas123 07 Jul 2010 07:55

read edited message above

Re: Recent Update commands used on ubuntu 9.10

samcargo 08 Jul 2010 06:04

Thank you again,

My 2TB WD crashed totally so i replaced it with a 1TB Samsung.
The Model i have fits only one Hard Disk not two.

Is this ok?

DISK_COUNT=0
WD_DISK_COUNT=0

let DISK_COUNT=DISK_COUNT+1
cat /sys/block/sda/device/model | grep WDC >/dev/null 2>&1
# if [ "$?" -eq "0" ]; then
let WD_DISK_COUNT=WD_DISK_COUNT+1
DISK1=sda
# fi

if [ "$MODEL" != "WWLXN" ]; then
let DISK_COUNT=DISK_COUNT+1
cat /sys/block/sdb/device/model | grep WDC >/dev/null 2>&1
if [ "$?" -eq "0" ]; then
let WD_DISK_COUNT=WD_DISK_COUNT+1
DISK2=sdb
fi
fi

echo -n "Checking WD disk number….."
if [ "${WD_DISK_COUNT}" -lt "1" ];then
echo "No disk found"
empty_conf
/usr/sbin/wixsendevent system factoryreset 1 >/dev/null 2>&1
stop_network
exit 1

Last edited on 08 Jul 2010 06:05 by samcargo Show more

Re: Recent Update commands used on ubuntu 9.10

tomas123 08 Jul 2010 07:39

I have not enough time entry deeper in your problem.
you post all outputs from your ubuntu PC - but your whitebook drive works fine - you have a web GUI !
make a ssh login to yout whitebook and post

1. dmesg
2. sfdisk -l
3. df
4. cat /proc/partitions
5. cat /proc/mounts

(use dmesg http://ifile.it/)

Do you tried a "factory Default" over the Web Frontend?

with a ssh login you can also start the factoryDefault.sh with the commandline.
then you get more informations

you are close to the target

Re: Recent Update commands used on ubuntu 9.10

tomas123 08 Jul 2010 09:18

there is another way:
http://highlevelbits.free.fr/index.php?option=com_content&view=article&id=107%3Amybook-rescue&catid=61%3Amybook-rescue&Itemid=78&lang=en

Re: Recent Update commands used on ubuntu 9.10

samcargo 08 Jul 2010 19:20

Dear Tomas,

Thank you very much! It worked finaly!!! I have done it all over using ubuntu 9.10
From the begining step by step.
Vielen Dank, Merci, Thank you!

Re: Recent Update commands used on ubuntu 9.10

tomas123 09 Jul 2010 07:41

Why it worked now?
What is the difference from the previous attempt?
It is important to me.

Re: Recent Update commands used on ubuntu 9.10

godovic 13 Jul 2010 08:20

Can anyone confirm that http://mybookworld.wikidot.com/install-clean-hdd-in-white-light-my-book-world-edition is working with newest firmware that can be downloaded - 01.01.16.img

Were can I download 01.00.16.img
Maybe I will get it to work with that firmware.

My problem is that after all steps with original disk I get only first white LED light, and the last one is blinking slowly.

Re: Recent Update commands used on ubuntu 9.10

tomas123 13 Jul 2010 08:31

sorry I don't read carefully your post
I only checked firmware 01.00.16.img

let's ask samcargo - he succesfull patched a firmware on 8 Jul 2010 (see above)

wget http://download.wdc.com/nas/wdhxnc-01.01.16.img -O ./fw.img

I think you are on the save side if you use the old firmware

http://support.wdc.com/download/mybook/WD-MyBookWorld-v1.00.16-GPL.tar.gz

then than make a firmware update with the GUI to V 01.01.16

Last edited on 13 Jul 2010 09:14 by tomas123 Show more

Re: Recent Update commands used on ubuntu 9.10

godovic 14 Jul 2010 20:09

I need to compile or? Doesn't have experience.. :(
Were can I find older firmware?

tnx for your time

Re: Recent Update commands used on ubuntu 9.10

tomas123 14 Jul 2010 21:54

http://support.wdc.com/download/mybook/WD-MyBookWorld-v1.00.16-GPL.tar.gz

Re: Recent Update commands used on ubuntu 9.10

godovic 15 Jul 2010 08:13

Ok… it is 430MB I will get it and maybe I will find how to get to img file…

Re: Recent Update commands used on ubuntu 9.10

tomas123 15 Jul 2010 08:51

- sorry, my mistake :-(
- the link above is only the source code
- the link to the old firmware image wdhxnc-01.00.16.img is down
- I think you can use the current firmware http://download.wdc.com/nas/wdhxnc-01.01.16.img

try it

Re: Recent Update commands used on ubuntu 9.10

tomas123 15 Jul 2010 20:03

ok, I uploaded it
http://mybookworld.wikidot.com/forum/t-253643/link-to-old-firmware-wdhxnc-01-00-16-img

Re: Recent Update commands used on ubuntu 9.10

godovic 15 Jul 2010 21:55

tnx a lot.. I will try to unbrick tomorrow morning or on Monday and I will write a feedback here.. wish me a luck :)

Re: Recent Update commands used on ubuntu 9.10

godovic 16 Jul 2010 10:05

nothing again… grrrr. :(

Step by step there is no error… The disk are mounted on the Ubuntu, the data is saved, but the disk is not recognized :(
There is 6 led, in 3 groups. The first LED is lighting and the lastone 6 is blinking. 5 seconds light than turn off and on for another 5 sec …. That all, I cannot reach the book via network interface.
I removed the disk, turn up again the NAS and the LEDs are lighting same way :(
The disk is working properly cos the Ubuntu is working with it normally and I double checked it with USB interface…

I changed the disk with Samsung 250GB just to try and edited the 158 and 161 lines with # - to disable IF FI statement but I am not sure, I have 1disk WD white, shoud I disable on the sda or on the sdb lines?

that's all … any suggestions? maybe the book is dead and cannot recognize any disk? the network interface is establishing connection but cannot even resolve IP, ping or …

should the network cable be connected for WD book to boot up ???

Last edited on 16 Jul 2010 10:07 by godovic Show more

Re: Recent Update commands used on ubuntu 9.10

tomas123 16 Jul 2010 10:15

I know, the tutorial works fine. You can read the success messages above from users gijsh and samcargo

a single disc white book is sda

samcargo above wrote

My 2TB WD crashed totally so i replaced it with a 1TB Samsung.
Thank you very much! It worked finaly!!! I have done it all over using ubuntu 9.10

repeat all steps carefully
the network cable is unimportant

Last edited on 16 Jul 2010 10:17 by tomas123 Show more

Re: Recent Update commands used on ubuntu 9.10

godovic 16 Jul 2010 10:36

How should be the LEDs light if the error is on HDD or on hardware..
Is there any status scheme 'cos there is 6 different LEDs?

Re: Recent Update commands used on ubuntu 9.10

tomas123 16 Jul 2010 11:02

the LEDs are switched by software from disk (no firmware inside on flash etc)

The first LED is lighting and the lastone 6 is blinking. 5 seconds light than turn off and on for another 5 sec

-> I think, your disk is booting fine, but your partitions are buggy
you can build a serial adapter and read the kernel panic message ;-)
see here http://mybookworld.wikidot.com/whitelight-serial

Last edited on 16 Jul 2010 11:06 by tomas123 Show more

Re: Recent Update commands used on ubuntu 9.10

godovic 16 Aug 2010 09:14

So.. yeap.. probably my partitions was buggy.

So I got new HDD WD WE 1TB white and disassembled the box and then clone the new disk to old one.
Now everything is fine.
So ppl, I do a successfully clone with a Clonezilla live CD - with disk to disk option.

'Cos my new disk is still outside of the box, can anyone suggest me a fast and good program to clone a disk to file, and then I can upload it so anyone can do a restore.

tnx

Re: Recent Update commands used on ubuntu 9.10

Erik66 23 Aug 2010 21:54

Hi there! I got the same problem as you! I used all the firm in the guide but it diddent work on my 1GB withe WD so there must be a poblem! All the steps in te guide workd without anny error! but the first light turn on and the last start blinking as you say!

so Is can you please send me a message On the link og a way to get your cloned file! cant you make a burn image form disk?

tnx

Re: Recent Update commands used on ubuntu 9.10

godovic 26 Aug 2010 13:45

I have not cloned my disk to file.. just disk to disk with my new WD WE. CloneZilla detect partition as RAW and clone every sector. So I guess, image file will be 1TB which is unacceptable. Thats way I asked here: I need recommendation for software that will clone disk to file without cloning empty space. The file will be smaller so I can upload it somewhere and help many ppl

Re: Recent Update commands used on ubuntu 9.10

tenshishi 12 Feb 2011 15:06

tomas123 15 Jul 2010, 16:03 GMT
ok, I uploaded it
http://mybookworld.wikidot.com/forum/t-253643/link-to-old-firmware-wdhxnc-01-00-16-img

Hi Tomas, I've tried with the latest firmware version, I've followed your step-by-step procedure more than once, but each time I plug the WD back, it never boot.

I'm asking if this is coming from the latest firmware version.
Could you please send me the one on your procedure ? I mean 01-00-16

As I actually fail each time I try with 01-02-06
wdhxnc-01.02.06_010411.img

No luck :(

boris198 31 Aug 2010 08:59

I use a Ubuntu 9.10 and follow the steps one by one.

I got and error on part 3 of the manual:

root@boris-desktop:/home/boris/Desktop/hd# cat /proc/mdstat
Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
unused devices: <none>

root@boris-desktop:/home/boris/Desktop/hd# ls /dev/m*
/dev/mcelog /dev/mem /dev/mixer

/dev/mapper:
control

root@boris-desktop:/home/boris/Desktop/hd# mdadm —zero-superblock /dev/sdb1
mdadm: Unrecognised md component device - /dev/sdb1

root@boris-desktop:/home/boris/Desktop/hd# fdisk -l /dev/sdb

Disk /dev/sdb: 250.1 GB, 250059350016 bytes
255 heads, 63 sectors/track, 30401 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x0001cf00

Device Boot Start End Blocks Id System
/dev/sdb1 5 248 1959930 fd Linux raid autodetect
/dev/sdb2 249 280 257040 fd Linux raid autodetect
/dev/sdb3 281 403 987997+ fd Linux raid autodetect
/dev/sdb4 404 121601 973522935 fd Linux raid autodetect

Can somebody assist me with it?

Re: No luck :(

tomas123 31 Aug 2010 12:22

you don't read the tutorial carefully:

"you need the package mdadm for creating software raids (go to System/Administration/SynapticPackageManager and install madadm - the tool to administer Linux MD arrays (software RAID))"

This installs also the device files.
You can create the device file by yourself

# mknod /dev/md0 b 9 0
# mknod /dev/md1 b 9 1
# mknod /dev/md2 b 9 2

Last edited on 31 Aug 2010 12:30 by tomas123 Show more

Re: No luck :(

boris198 31 Aug 2010 14:37

Thank you for replay.

Actually i read and install the mdadm (manually, the pc that i work on it did not have an internet connection).

I already create the device files, but i still did not succeed:
I got the "mdadm: Unrecognised md component device - /dev/sdb1" on mdadm —zero-superblock /dev/sdb1

I can continue and finish the procedure, but the MyBook did not booting (only top led shining)

Probably will no longer work

mikeywikey 12 Feb 2011 18:11

Just got an empty white light box from eBay £22 + postage (seems a fair price), as I run a few of the blue rings I'm reasonably familiar with this sort of kit (and I'm a UNIX admin), it looks like there was a major version update at 01.01.18 which changed all the offsets (see upgrade1.sh).

"part2: build the boot partition" will probably need amending with the new offsets (unless you can find an old 01.01.16 upgrade file, use that then upgrade officially).
I'll put the new offsets here (plus the additional step), once I've checked they work.

Good guide by the way :-) saved me lots of hacking!

Re: Probably will no longer work

alllexx 12 Feb 2011 18:40

That probably means they rebuilt the u-boot binary with different settings. For what purpose, I wonder? How strange of them…

Re: Probably will no longer work

mikeywikey 13 Feb 2011 14:42

My guess is that they are preparing for the >2Tb drives, you have to move from MBR to GPT (2Tb is the limit for MBR partitions) and that might needs some extra space, look like they have shifted it by the size of the GPT stuff - Just a guess, but it would be interesting to run a white light with a 3Tb drive, format it up as GPT and see if it works.

Re: Probably will no longer work

alllexx 13 Feb 2011 16:50

My guess is that they are preparing for the >2Tb drives

Sounds reasonable.

it would be interesting to run a white light with a 3Tb drive, format it up as GPT and see if it works.

I'd advise against doing this: MBR contains the first stage bootloader that launches u-boot, so unless you patch GPT to launch the u-boot binary at ${UBOOT_START}, it obviously won't work.

UPD: my bad, of course, MBR just loads the stage1 bootloader, which in turn loads das u-boot :)
So, you'll have to patch GPT to load the stage1 binary at ${STAGE1_START}

Last edited on 13 Feb 2011 16:59 by alllexx Show more

Re: Probably will no longer work

mikeywikey 13 Feb 2011 17:17

I suspect that it understands GPT to some extent, or at least the new kernel does becuase when I didn't zap the secondary GPT (end of the disk) it tried to add a GPT partition as part of the factory initialse which toasted the disk into a single ee MBR partition, I didn't mean to say that simply formatting a drive as GPT might work, obviously there's a little more to it, but getting stage1 to kick in would be the key, I suspect that we *might* see 3Tb and 6Tb MyBooks for sale sometime soon (which will give us the upgrade path), it all depends if it will be based on new hardware, or just firmware (I don't think there's any chance of blue rings supporting the big drives without some hacking as the kernel doesn't support large filesystems >2Tb).

Re: Probably will no longer work

alllexx 13 Feb 2011 18:09

I don't know about the oxnas_810_eabi_dse_defconfig (haven't tried it), but when I was struggling to get whitelight's kernel to work with bluerings (the kernel claims to provide ox800 platform support, builds for this target after some patching and is even successfully booted by u-boot (I can tell, since it doesn't switch to booting from the second drive), but doesn't go any further: leds don't light on, so no kernel modules are loaded) I noticed that GPT support is available in the kernel config. It all boils down to simply having this option enabled (if not already) and stage1 bootloader being launched by GPT, so no new hardware is needed, imho.

Last edited on 13 Feb 2011 18:09 by alllexx Show more

tenshishi 12 Feb 2011 19:23

I m trying to revive my My book world edition white light 1Tb model with a WD caviar green 2Tb HDD.

I tryed following all the howto i found and this wiki, but nothing worked.

I also tryed the HDD clean install for white light many times.
But it seems as mikeywikey said, that there were major updates on new firmware avalaible from WD site. offsets should have changed.

Actually I only found the new firmware 01.02-06 ( http://download.wdc.com/nas/wdhxnc-01.02.06_010411.img )
I also noticed that the procedure was done with an older firmware that I was unable to find on the net.

( I also paid on Paypal for the MyBookRescue (with mybookformat and mybook restore).
Tryed everything.. Didn't work.

I Just want to recover to WD original features.

Could someone help me installing new 2Tb WD HDD in a WD World Edition 1Tb model White Light

Thanks for you help.

Last edited on 12 Feb 2011 19:25 by tenshishi Show more

New offsets

mikeywikey 13 Feb 2011 15:03

Okey dokey….

This is what I did (everything assumes /dev/sdb - so be careful if it's not!)

1. Follow the instructions in the guide to unpack and decode the download from WD

2. Initialise the disk (note, this will completely clean it - you will lose all data)

# Wipe all partition tables (including GPT if present)
# Start of disk
dd if=/dev/zero of=/dev/sdb bs=512 count=64260
# End of disk (the number is the last sector, found by cylinders x cylinder size (my disk was a WD 2Tb EARS)
dd if=/dev/zero of=/dev/sdb bs=512 seek=3907024065

# Put the partition table on, actually, I'm just being lazy, you need the MSDOS header
dd if=./fw/mbr of=/dev/sdb bs=512 seek=0

# Create all partitions with the correct size - sdb4 resized to max
(
echo 4,244,fd
echo 248,32,fd
echo 280,123,fd
echo 403,,fd
) | /sbin/sfdisk --force /dev/sdb

3. This is the new part2 with the new offsets (replaces part 2 of the guide)

BS=512

STAGE1_START=36
UBOOT_START=38
UPGRADE_FLAG_OFFSET=290
KERNEL_START=336
INITRD_START=16674
INITK_START=8482
STAGE1_BACKUP_START=32178
UBOOT_BACKUP_START=32180
KERNEL_BACKUP_START=32478

MBR1st=fw/1stMBR
STAGE1=fw/stage1.wrapped
INITRD=fw/uUpgradeRootfs
INITK=fw/uImage.1
UBOOT=fw/u-boot.wrapped
KERNEL=fw/uImage

# backup stage-1 image
    dd "if=${STAGE1}" of=/dev/sdb bs=${BS} seek=${STAGE1_BACKUP_START}
# backup u-boot image
    dd "if=${UBOOT}" of=/dev/sdb bs=${BS} seek=${UBOOT_BACKUP_START}
# backup kernel image
    dd "if=${KERNEL}" of=/dev/sdb bs=${BS} seek=${KERNEL_BACKUP_START}
# main stage-1 image
    dd "if=${STAGE1}" of=/dev/sdb bs=${BS} seek=${STAGE1_START}

#update 1st boot info in MBR if version is before 01.01.18 this is vital for an upgrade (not needed for this as it will have this in already, but doesn't hurt, I left it in here as it was the key to the version)
    dd "if=${MBR1st}" of=/dev/sdb bs=1 seek=432

# main u-boot image
    dd "if=${UBOOT}" of=/dev/sdb bs=${BS} seek=${UBOOT_START}
# backup kernel image
    dd "if=${KERNEL}" of=/dev/sdb bs=${BS} seek=${KERNEL_START}
# main upgrade rootfs
    dd "if=${INITRD}" of=/dev/sdb bs=${BS} seek=${INITRD_START}

# main upgrade kernel
    dd "if=${INITK}" of=/dev/sdb bs=${BS} seek=${INITK_START}

4. Build filesystems

This is part3 - you can do it as documented, you don't need the last section about correcting the MBR as we have done it already

Hope that helps

Last edited on 13 Feb 2011 15:06 by mikeywikey Show more

Re: New offsets

alllexx 13 Feb 2011 18:56

Your new part 3 is missing setting u-boot environmental variables. If nothing relevant is mentioned in the upgrade1.sh script, it means the offset for them hasn't changed, so just add it from the original howto. I'm not sure whether it affects anything apart from mac-address, but better be safe.

Also, I have slightly patched u-boot's fw_printenv utility to be able to change u-boot's environmental variables on bluerings, and I think it can be used with whitelight as well. I've built an armel binary, but you may also need to edit /etc/fw_env.config
If you're willing to try it, follow this howto, Method 1, (don't forget to write env variables beforehand) and just use uboot-env.v2.armel.tar.gz instead of uboot-env.v2.arm.tar.gz.

UPD: there's one thing that bothers me: if, like in bluerings the address, from which kernel image is loaded is set in u-boot's environmental variables, they have to somehow change them according to the new offset. If there's no sign of them doing this in upgrade1.sh, I wonder if they do this in the upgrade initrd?

Last edited on 13 Feb 2011 19:29 by alllexx Show more

Re: New offsets

mikeywikey 14 Feb 2011 03:46

If you don't set a mac you'll get the default Oxford Semiconductors one (00-30-E0-00-00-01) this isn't a problem unless you have more than one white light box on the same IP segment, almost certainly the location has changed (although I havent given it any thought), when I get some time to play I'll have a look - thanks for the link.

Re: New offsets

alllexx 14 Feb 2011 08:10

If you don't set a mac you'll get the default Oxford Semiconductors one (00-30-E0-00-00-01)

Yes, I'm aware of that, furthermore, all the u-boot settings are set to default if u-boot fails to varify the crc32 checksum of the existing env variables.

almost certainly the location has changed

It would be interesting to verify this.

AFAIK, guys at WD have never before somehow changed u-boot's env variables, since there're individual for every single box and you can't change them without calculating and updating the checksum. I'm just bothered how they managed to alter them without invalidating the checksum. And if they haven't, how in the world do they boot kernel from the new address?

thanks for the link

You're welcome

Last edited on 14 Feb 2011 08:12 by alllexx Show more

Re: New offsets

sorinacke 15 Feb 2011 21:34

Dear Mickey,

if a have 2 working drives and i want to upgrade both of them to a bigger one, how do i copy the MAC address part from old drive to the new one?

Using tomas addresses i should do on the old HDD

1. main MAC image

dd if=/dev/sdb of=./fw/mac bs=512 count=2 ?? for 945 bytes?
and get the file on the HDD

and write in your script with a 36 bytes shift on the new HDD

STAGE1_START=36
MAC_START=274
MAC=fw/mac

1. MAC image

dd IF=./fw/mac of=/dev/sdb bs=512 seek=239

unfortunately it does not work. i am missing something (except for the Linux Knowledge :).)

Last edited on 16 Feb 2011 09:25 by sorinacke Show more

I can help

sorinacke 18 Feb 2011 18:51

i have finished recently. just pm if you need help (still)
S

Re: I can help

tomas123 18 Feb 2011 20:17

you post some news in many parts
please collect your experience and hints and post them here (at the end of site)

is create and format /dev/md4 the solution?

Last edited on 18 Feb 2011 20:17 by tomas123 Show more

Re: I can help

sorinacke 20 Feb 2011 21:03

I have identified 2 problems:
1. with MAC address which i solved using a solution to change the MAC after startup of the device, with a script. I posted the solution here for Mickey to help me get the MAC from the original disks.

2. the error at the end of the site with 0.90 superblock …. format /dev/md4 did not help.
Still not working. It was solved temporaroly and then later after reboot, i have the same error message.

I built the second hdd and it worked the same: sda4 does not have a valid v0.90 superblock, not importing!

The HDD is accessible, and can read/write data. i do no tknow is it is ok to start using it with this error.

To save other guys from investing such a long time like me, i will post the full script used, but also a link to the working software.
Tomas, if you have saved the working version somewhere, then we might have not need new offsets and stuff when WD changed the version. you install a working version start HDD and then hit update firmware. Job done.

Last edited on 20 Feb 2011 22:28 by sorinacke Show more

Re: I can help

mikeywikey 21 Feb 2011 11:52

Hiya,
Been a bit busy so haven't has a chance to reply, looks like the previous problems installing were because of a couple of issues;

1. Watch your blocksize, the guide assumes 512 bytes, however some new disks (specifically, the larger advanced format disks) will come with 1024 bytes or even GPT style 4094 bytes (from the log I see the disk was a 1024 byte).

2. If you build your arrays under ubuntu 10.10 (possibly anything with a 2.6 kernel) then the md superblock can be later than 0.90 (1.x, or 0.91 in some circumstances), this depends on your mdadm.conf, but forcing -e 0.90 will ensure an older superblock - you would not get this issue under the version of unbuntu in the guide.

Changing the ethernet address via a boot script is probably the easiest way (and quite tidy).

tomas123 14 Feb 2011 11:24

@mikeywikey
great work, I changed my post http://mybookworld.wikidot.com/install-clean-hdd-in-white-light-my-book-world-edition
If you want to edit this page - please extend the tutorial - don't remove the part2 for old firmware update

I got many PM about this problem with fw above 01.01.18
Why nobody noticed my hint

# for infos about the DISK SECTOR OFFSETS see ./fw/upgrade1.sh

Last edited on 14 Feb 2011 11:26 by tomas123 Show more

sorinacke 15 Feb 2011 19:26

I have 2 pieces, original with a 1TB WD caviar Green. I want to upgrade to a 2TB WD Green Power disk.

I have run the upgrade as described by tomas123 with no notable error.

When i connect the electronic board, led 1 on, led 6 on and off then stays off. Led on network only green blinking regularly. Yellow one stays off. No IP is takes, it does not boot!

1. sorry to post the log, but i have no idea what can be.
2. If i have 2 pieces and upgrade them will they get the same MAC address? Does it bother in the same network?

Complete log can be found here

http://pastebin.com/tNZ0Q9kz

Thank you!

Last edited on 15 Feb 2011 20:16 by sorinacke Show more

tomas123 15 Feb 2011 19:48

@sorinacke
Please shorten the text!
post your code to http://pastebin.com
Thank you

sorinacke 15 Feb 2011 20:00

Done. I have not figured out how to insert in here in a box, just copy/paste the link ….

Completions

sorinacke 15 Feb 2011 19:51

Completions:
1. i tried hard before asking for help. I am trying since almost 1 week to fix it and i did not succeeded. My Linux knowledge is not as good as it should maybe
2. i have 2 working devices, original so i could copy parte of the disks. i am afraid to touch them not to loose data, but MAC address section or other setting might be useful …

3. Teamviewer remote access is a friend of mine :). This way we can check again head to toe the tutorial with the latest firmware to date. So if you have time / mood / just let me know

Finally i have it online (thank to the Mickey guy, long live ….) with the following error in dmesg
…….
md: Autodetecting RAID arrays.
md: invalid raid superblock magic on sda4
md: sda4 does not have a valid v0.90 superblock, not importing!
md: Scanned 4 and added 3 devices.
….

the rest of it is like in the original :).

OS: Ubuntu 10.10 running in Virtual machine - wmware.

Last edited on 20 Feb 2011 20:55 by sorinacke Show more

Mickey - thank you for tips

sorinacke 21 Feb 2011 21:30

Dear Mickeywikey,

Thank you for the lines dropped above.

I installed the disks anyway and until now it seems it works.

2. What if the tutorial was for 512 blocks and i have 1024 blocks? are the starting addresses incorrect? Can i format the disks with smaller blocks - like in windows? I see not other disadvantage but wasting some space for the system files using bigger clusters. For the data volume anyway, considering the size of the files (at least 100MB one) - it does not matter.

3. does it make sense to take care of the 0.90 error? why does it appear only on sda4? all others partitions were done in the same way….

I do not have any experience with the Software RAID - maybe the questions above sound strange.

Last edited on 21 Feb 2011 23:14 by sorinacke Show more

page 1 of 212next »