New Release - Firmware Version 2.3.01 for the WD Sharespace (9/12/11)

Community Archives WD ShareSpace
41

aKazak - I gave up for today. It’s just taking me too much time. However - it’s not the end - planning to get back to this on tomorow. In the mean time, just some thoughts: - YES - serial cable would be an option, but I dont’ feel like building it up, for me is enought to take the hdds out of it, and conect them to big box all th etime I want to edit something. :slight_smile: And I am definitelly not planning to play with this nas in the future as much - i just need good fw. Already I learned that all you really need is to take backup of / before fw update…

Something for all - do you know the meaning of third partitions of your SS? Mine are the same size as first (214MB), but contains nothing, just occupiing space. I thought maybe this will/should be backup of previous version of system, but - as I said - on my box, it contains just nothing… (first 214MB partition is root, second is swap, third 214MB is this “backup”, containing nothing, and 4th is data)

So for tomorow - I am planning to get the disks out, connect them and

  1. take a copy of / partition

  2. check the size and if big, get rid of /proto/SxM_webui/admin/config.xml

  3. check other logs to try to catch some error.

  4. edit /etc/init.d/rcS to start just network and ssh daemon

  5. put it back to box and try to boot.

  6. post result on this forum

If you have better ideas - please let me know within next 24 hours.

thak you all

42

Jannem - stay calm, unless you issued factory restore (which writes zeroes to your data partition), your data should be still ok on last partitions of you disks, regadless of fw updates - well at least in my case (8TB box, upgraded to 2.3.01) the raid5 data were ok.

The raid , hdd partitioning, filesystem used in wd box is very same as in most other linux distos (because this also runs linux). There is no hardware raid.

So there is a “simple” way to get access to your data somewhere else (assuming we are not going to fix the box now, just to get the data out of it).

So what you need to do:

  1. get all 4 drives out of SS (mark them 1-4 to know later their correct positions)

  2. connect them to some linux box and get some info about hte disks (dmesg info is enough -you just need to know what device your distro used for aech hdd - sdX - ie sdb, sdb…), you also may check fdisk -l /dev/sdX, to get the idea of your partitions on SS disks.

  3. assemble raid from 4 data partitions (do not create, just assemble) of 4 connected hdds (or their dd copies). The data partition should be the last = biggest one (sdX4).

so - something like this: mdadm --assemble /dev/md0 /dev/sdb4 /dev/sdc4 /dev/sdd4 /dev/sde4

To get idea - in my case I was not working directly with disks, but rather with their dd copies: mdadm --assemble /dev/md0 /dev/loop1 /dev/loop2 /dev/loop3 /dev/loop4 - I had dd copies of sdX1 SS partitions mounted in loops: (losetup /dev/loop1 /tmp/mmd1sdb1; losetup /dev/loop2 /tmp/mmd2sdb1; losetup /dev/loop3 /tmp/mmd3sdb1; losetup /dev/loop4 /tmp/mmd4sdb1).

If the raid is assembled, just mount it: mkdir /mnt/SSraid; mount /dev/md0 /mnt/SSraid

  1. your data should be in /mnt/SSraid - so just get them :slight_smile:

Let us know, if this helped.

43

Ok AKazak, I have done that restet by Erasing the 4 (1GB) HDDs ihave and I re-put them back and just put the power plug and the LEDs are turned on, the power button is blinking does that mean that the process of RAID management is still working or does it mean that I have to power on the device? actually i am afraid to press the power button now after suffering 48 hours of **bleep**… I wanna make it right this time.

44

aKazak:

After quick erase of 4 HDD’s mark each drive with number from 1 to 4 and insert to ShareSpace according to the numbers. Power on ShareSpace and wait 30-40 hours before factory restore is completed. In the result you will get your own clean md0 volume and you can easily back it up.

aKazak - from where should the system restore, when you erase (write zeros) to system partition?

I need to narrow down my point of view. → I think SS does not have any other place to save/run/restore system than md0, so I would not zero it out, unless I would have a backup copy of it.

But maybe you know from where the system will restore - that’s why/what I am asking.

Thank you for explanation.

45

there is a firmware on the SS device that runs all the operations and it manage the HDD RAID configurations even if you remove the HDD

46

OK - this i really a new an important infomration for me.

Thank you.

SO - my approach to fix this will probably change.

Moataz - were you successfull?

47

the blinking is stopped … so i think it finished. I am going to text it in a while. will tell you the result

48

OK - another small victory  - I HAVE SSH !!!

Steps:

  1. After connecting hdds to my linux box and mounting them; I did not find any big files after factory restore on md0 (i.e. /proto/SxM_webui/admin/config.xml is now 5kB), free space on md0 is about 78MB

  2. no relevant info in logs - but apparently I just don’t know what to look for.

  3. ssh daemon can by started by editing /etc/inittab:

add row 31 :“::sysinit:/usr/sbin/sshd”

so it looks like this:

now run any rc scripts

System startup

::sysinit:/usr/sbin/sshd
::sysinit:/etc/init.d/rcS
4. I will continue investigating this, so far I am too happy with ssh, I am almost forgetting the main reason for this - I am still unable to finish the initial wizard (system busy error aftre hiting finish in gui).

Any suggestions will be very much welcommed.

NEW and this time BIG VICTORY: after setting correct time in the device and resetting it  my initial setup went through!!!

 Well - currently I am not sure what accually helped (I am a little suspicious that it wan’t date/time issue - why would it be…); it is starting to turn out my hdd1 has badsectors, so maybe after some restart the box had put  this failing disk out of raid and let me save configuration on that remaining 3 disks; however - the inittab trick for enabling ssh worked (now I edited back innittab, as I am able to start ssh using gui)

If something reasonable turns out of this, I’ll let you know.

So - to summarize - after fw update - if your web service is not starting - try to allocate some free space on the box by using twonky (find box’s ip , twonky runs on port 9000, erase twonky logs)

If you still cannot start web service, you’ll probably have to take disk (acctually 1 disk should be enough) out and mount first partition to some linux box, to edit it (helpfull commands : “mdadm --asseble /dev/md0 /dev/sdb1 --run” mount /dev/md0 /mnt/some_name).

Check how much space is free (df -h, or du -hs).

If you need ssh, edit /mnt/some_name/etc/inittab.

If you want to force the system to make the factory restore (which may also help) you can touch /etc/.factory_restore

I did not zero-out my disks (as aKazak recommended), but this also might help you.

!!! be carefull about not editing last partitions on your disks (sdb4 in my case), if you need data!!!

49

OK. I’m a WD SS 4Tb user (on iMax OSx Lion) and since FW 2.3.0.1 my admin-configuration screen is only blank.

So the HTTP-service is not running. Also connecting by SSH is not possible (connection refused-error).

I already rebooted the device a zillion-times and also did a reset of it.

Twonky server is available and data is available through Finder. Drive is not recognized by Time Machine.

Issued a ticket to WD since 20/09, but no response…

Any suggestions what to do to be able again to enter the configuration-page of my SS?

All help is appreciated!

Best regards,

R.

50

AKAZAK’s way worked fine for me… :slight_smile:  thanx guys

51

Hi citronic,

Thanks for your advice. I am not really familiar with with the kind of command-line work you propose, but I will make some study of it before trying anything rash. You (and some others) talk about connecting the HD’s to ‘some Linx box’. Does this mean another NAS (barebone system with 4 empty bays)? Or does it mean a machine running Linux? I guess it’s the first option. I would then have to go out and buy a barebone NAS. Any advice on a good brand? Don’t bother making jokes about WD :wink:

So, if I understand correctly:

  • remove HD 1-4 from WD SS;

  • label HD’s 1 to 4

  • connect to a NAS box running Linux;

  • assemble raid from the 4 data partitions;

  • backup data from biggest data partition;

  • put back HD 1-4 in WD SS;

  • perform full factory restore;

  • put back data from backup to WD SS (if preferred).

I would very much appreciate any elaboration on handling the disks in connecting them to another system and performing those Linux-commands on them.

I run Mac OS X 10.7 (Lion) on a MacBook Pro. I have Parallels installed, but no Linux VM’s (as of yet). I also have a USB device to interface a bare SATA HD to my MacBook Pro via its USB port. If there are any solutions hidden in this configuration, I would love to hear it. Many thanks in advance.

Jannem

52

@citronik I have plenty of free space on my root partition…however, after doing what you suggested, I see a /old?  Can I copy this back & reboot?

Also, I can access twonky no problem & I have tried manually restarting the httpd service, nothing.

The logs do not tell me anything, this is the only thing they say: “GET /favicon.ico HTTP/1.1” 404

So, it looks like it httpd is running, but the pages are not rendering.

It almost looks like it is a problem with either the httpd config or php.

/ $ df -h

Filesystem                Size      Used Available Use% Mounted on

/dev/md0                  2.1M      2.1M         0 100% /old

/dev/md0                197.4M    102.9M     84.3M  55% /

/dev/vg0/lv0              1.8T    516.9G      1.3T  28% /DataVolume

/dev/ram0                61.6M    140.0k     61.5M   0% /mnt/ram/

53

Jannem - you can connect the disks to ANY Linux machine. So all you need is a computer with four sata ports, running linux.

Even the linux is not a must - I just tried it the linux way (Ubuntu btw, but it really doesn’t matter, as long as your distro will have packages for raid installed - vast majority have them) - beceuse SS box runs linux and the data structure is linux (raid + partitioning).

Except this fact (that you can use any linux pc) -you wrote a way which I would go.

I have to mention aKazaks way (also recommended by Moataz) of rebuilding the SS nas (!!!after you get the data out of it !!!) - that is - to write zeroes to first and last milion sectors of the SS disks = acctually to destroy any system data and let the SS box to recreate your system + CLEAN data partitions afterwards.

You have my full storry posted within previous pages of current thread; also you can get some idea about handling disks here (! be carefull not to follow exactly, because Nathan_H describes raid1; but later within that thread there are more comments about raid5): http://community.wdc.com/t5/WD-ShareSpace/HOWTO-Recover-files-from-a-RAID-1-mirrored-drive-when-the/td-p/138448

as I said before - dont’ worry, your data should still be there.

54

AppleFanboi - currently I am playing with through tests of my failing hd0 from the affected box.

I can’t check the box directly now, but from saved image of working box I see, that there is directory /old, but does not contain anything - so it’s probably only mounted on running box with some options (which cause showing only 2.1 MB occupied space on your box) - you can check this mount on your SS box by issuing “mount”. I would not copy it over to /

You can check what ports are open (which can give you idea of running services) by issuing “nmap ipaddress” on some linux box connected to same network as your SS.

You can check directly on your SS (I hope I remember good, that you have ssh access), by issuing " ps -ef | grep http" if the http service is running.

You can start httpd manually by issuing “/etc/init.d/S55mini_httpd start”

You should then check log file of http on your box: /var/log/mini_httpd.log

BTW: I am a linux user, because I used to work as a PC tech. :wink:

55

citronik wrote:

 

BTW: I am a linux user, because I used to work as a PC tech. :wink:

 

LOL, M$ is really good at doing that!

Now, regarding http:

HTTP is running, confirmed with the nmap from my linux server. 

Like I said, the logs don’t give me any info.  

I’m actually just tempted to leave it like this, at least I can connect to it & get my data.

I’ve already lost everything once, don’t want to go through that again.

Not shown: 1669 closed ports

PORT     STATE SERVICE

21/tcp   open  ftp

22/tcp   open  ssh

80/tcp   open  http

111/tcp  open  rpcbind

139/tcp  open  netbios-ssn

445/tcp  open  microsoft-ds

548/tcp  open  afpovertcp

802/tcp  open  unknown

1025/tcp open  NFS-or-IIS

2049/tcp open  nfs

3689/tcp open  rendezvous

Nmap finished: 1 IP address (1 host up) scanned in 0.212 seconds

56

AppleFanboi - I would really like to know, what is going on on your SS, when you issue in ssh:

“ps -ef | grep http” - how many http processes are there running?

I am not 100 percent sure, if /etc/init./S55… script will work, so lets try the directone:

" /usr/sbin/mini_httpd stop" - does http service stop after this (you can check with nmap)?

" /usr/sbin/mini_httpd start" - does http service start after this (you can check with nmap)?

And maybe - is anything showing up in browser when you issue some real php address that should work on your box:

“http://yourIPaddress/admin/system\_index.php?lang=en”

?

57

~ $ ps -ef | grep http > 2293 root        572 S < /usr/sbin/mini_httpd -d /proto/SxM_webui -c **.php|**> 18494 root       580 S   grep http 

Stopping the service - checked with nmap, port 80 closed

Started the service - checked with nmap, port 80 open

As I said before, I suspect it has to do with php since the https service seems to be running, but the pages are not rendering.

58

Hmm I am afraid I don’t have any advices for you now. 

I was probably not paying enough attention to your previous posts (my fault - don’t remember you already suspecting php) - sorry.

As for http - it seems to be doing what it should.

You are right - the php must be somehow screwed (I wonder if there is some code in /proto… don’t know the exact route in your SS box)

So - for now - I can’t help you. Maybe, after I turn my SS box back on (planning to do it tomorow) I’ll run to something useful for you.

Peace.

59

Thanks though,

hey at least we got to try a few things.

60

My bad experience of restoring data by means of Linux box

After sucessfull Factory Default Restore (with prior data backup and Quick Erase of 4 HDDs) I decided to copy the data from NASDRIVE back using Linux box. The main reason is time: SATA (even with RAID 5) is much faster than Gigabit Ethernet. After restore (copy -a … … ) finished I inserted all 4 HDDs in the correct order to the ShareSpace and powered it on. The result - all 4 HDDs LEDs become orange, HTTP user interface tells than all 4 HDDs are “failed” and the only option is to rebuild RAID array from the scratch.

Does any of you faced this problem before? I clearly remember that running version 2.2.91 of firmware I did the Linux restore trick several times!