I guys, i need help, i only have 100MB on the raid disk system, i need to increase.
So i want or convert to linear or resize the raid.
Please help.
Hi. I recently purchased a WD MyCloud 3TB because it looked like the perfect low-powered linux NAS.
First thing was certain, it came with v4.00 firmware out-of-the-box. Downgraded to v3, all was good.
Except I couldn’t install some software because of version conflicts with libc6.
I came across this thread and the prospect of using a less restrictive Debian installation seemed reasonable. I followed the guide on page 1 to installing the OMV 1.17 image with Debian. Everything seemed to work no dramas on first reboot.
Ran the web interface and did some basic configuration and update (I did not run the post-install script however). I changed the hostname of the device from “MyCloud” to “djtubig-mycloud” without quotes. Seemed OK so far.
Because I didn’t run the post-install script, I changed the locale manually and deleted the russian locales (because I don’t understand it)
Typed reboot, SSH disconnected and now while the Mycloud LED is green, my router is no longer picking it up. The ethernet on the back of the Mycloud only flashes one light; the amber light does not turn on. I’ve tried pulling the power plug and back in. I’ve tried using the pinhole reset button. No dice.
What did I do wrong here? The post-install script was optional so I didn’t run it. Is my device now soft-bricked for changing the hostname? It looks like it’s turning on but seems to fail to initialize the ethernet on reboot. Hard to tell. 
Yeah, few mans told about this error, but i cant get it (All works fine).
For now - yes, you have device with brocken network setup.
Recovery is easy - need edit /etc/network/interfaces manually (Connect HDD direct to PC or use “Magic” recovery package
What is this “Magic Recovery Package” you speak of? I’m interested in trying to fix the configuration, but I need a way to access the disk (preferably without breaking apart the box). I saw a thread relating to communicating with the device before boot using a magic packet, but the instructions are incomplete (i don’t see a “package” as far as I can tell).
More information would be greatly appreciated.
MaliceX wrote:
Hi. I recently purchased a WD MyCloud 3TB because it looked like the perfect low-powered linux NAS.
First thing was certain, it came with v4.00 firmware out-of-the-box. Downgraded to v3, all was good.
Except I couldn’t install some software because of version conflicts with libc6.
I came across this thread and the prospect of using a less restrictive Debian installation seemed reasonable. I followed the guide on page 1 to installing the OMV 1.17 image with Debian. Everything seemed to work no dramas on first reboot.
Ran the web interface and did some basic configuration and update (I did not run the post-install script however). I changed the hostname of the device from “MyCloud” to “djtubig-mycloud” without quotes. Seemed OK so far.
Because I didn’t run the post-install script, I changed the locale manually and deleted the russian locales (because I don’t understand it)
Typed reboot, SSH disconnected and now while the Mycloud LED is green, my router is no longer picking it up. The ethernet on the back of the Mycloud only flashes one light; the amber light does not turn on. I’ve tried pulling the power plug and back in. I’ve tried using the pinhole reset button. No dice.
What did I do wrong here? The post-install script was optional so I didn’t run it. Is my device now soft-bricked for changing the hostname? It looks like it’s turning on but seems to fail to initialize the ethernet on reboot. Hard to tell.
MaliceX,
I got exact same issue when I started to use OMV on WD. It tooke me half a day to figure out the root cause.
When I changed hostname or domain name from web interface, my network configuration for eth0 was gone, so no IP for my NIC any more. I had to open the box to re-flash…
auto eth0
allow-hotplug eth0
iface eth0 inet dhcp
@Fox_exe: I should apologise, I didn’t realise the recovery package on your post was a linkl. I just used that and following the instructions provided, I was able to communicate with the Mycloud device via Telnet.
Interestingly, having checked /dev/md126 which was the original filesystem, based on @j3prof’s note that the eth0 entry was removed. I put the entry back into /etc/network/interfaces in the respective mounted filesystem, and now when booting, the MyCloud LED remains white on normal boot.
Still not connecting to my LAN though.
EDIT: I’ve attached my /var/logs/messages file as-is. Note that Apr 7 is actually 26th April 2015 (I didn’t fix this at the time yet haha).
https://dl.dropboxusercontent.com/u/1287967/messages.log (Note: Windows line endings)
I guys, i need help, i only have 100MB on the raid disk system, i need to increase.
So i want or convert to linear or resize the raid.
Please help.
@Fox_exe: I ended up doing an image restore. With some help from your guide and that “Magic” recov package you linked and by using the Telnet method (which looks to be similiar @baerle’s approach of probing the device on boot), I have successfully unbricked my WD My cloud 3TB without removing the hard disk at all or opening up the enclosure!
Had to use ftpd to upload the images onto the disk after mounting /dev/sda4 (the main data partition), force-recreate the md0 RAID partition with mdadm (step 8-1 on your restore guide), then using dd as normal (keeping careful note of the correct /dev/sda# partition numbers).
I’ll probably give OMV another go at a later date, but I currently don’t have time to toy with it further, in fears of screwing up configuration once again. But it’s really good to know unbricking/firmware restore can be done without physical destruction now!
@meirinha: I think you’ll find that to do that, would mean you would need to do the following:
- Decrease /dev/sda4 partition size to allow for the extra space, also realigning further away from /dev/sda3
- Re-align partitions /dev/sda2 and /dev/sda3 forward
- THEN you can increase /dev/sda1 and /dev/sda2 partition size.
I would think this would only be possible to do either by removing the HDD and managing the partitions on your desktop machine, or theoretically it could be possible using the Telnet method. Chances are, you will need to lose your data to be able to do all this correctly.
Then there’s the issue of writing rootfs.img onto the resized /dev/sda1 and /dev/sda2 which is where your OS partition sits. Since most rootfs.img files are 2GB accounting for the 2x1GB RAID distribution, you’re going to have a hard time recovering your images this way.
Long story short, it’s possible but it will require a lot of effort.
To MaliceX:
Need rebuild mdraid to normal state:
# Check md settings:
cat /proc/mdstat# You get something like this:md127 : active raid1 sda1[0] sda2[1] 1999808 blocks [2/2] [UU]
# if you have only one md (md127) with both partitions (sda1/sda2):
mdadm -S /dev/md127
mdadm -A /dev/md0 /dev/sda1 /dev/sda2
mount /dev/md0 /mnt
nano /mnt/etc/network/interfaces
# edit it and save.
# unmount all and reboot:
umount /dev/md*
sync
reboot -f
But if you have 2 md’s (md127 and md128 or 126):
cat /proc/mdstat
# You get something like this:
md127 : active raid1 sda1[0]
1999808 blocks [2/1] [U_]
md128 : active raid1 sda2[1]
1999808 blocks [2/1] [_U]
# Mount both and find correct partition (If folder /etc/openmediavault/ exist - its OMV. If you see /etc/wdcomp.d - its official)
mount /dev/md127 /mnt
ls /etc
umount /dev/md127
# Next step - find what md number used for boot:
dd if=/dev/sda7 bs=1k count=1
# Search string like "root=/dev/mdX" (where X is 0 or 1)
# For example - omv on sda1 (md127) and we need md0 for boot.
mdadm -S /dev/md127
mdadm -A /dev/md0 /dev/sda1
mount /dev/md0 /mnt
nano /mnt/etc/network/interfaces
# Edit and save
umount /dev/md0
sync
reboot -f
To meirinha :
You need clean apt cache (With WebGUI or via SSH command: apt-get clean
Also need move DLNA cache (Replace [Tab] to disk UUID or just press [Tab] button when type paths):
service minidlna stop
mv /var/lib/minidlna /media/[Tab]/
ln -s /media/[Tab]/minidlna/ /var/lib/
service minidlna start
* Same for Plex. (Replace “minidlna” to “plexmediaserver”)
And dont use temp folders for Transmission! (Uncomplete dir enabled = false/uncheck)
Yeah, I had md126 and md127 appearing on my box. This came a few hours too late I’m afraid, but I’ll be sure to investigate that in future since I’ve started from scratch on v3 official. 
Seriously though, many thanks for your work on MyCloud and your written guides. I’ll definitely try again at a later date just so I can make sense of why it’s happened.
HI Guys,
can you help me with my FTP Failure ?
Can you see under:
Hi all,
Like many of you, i was mad at the stock system, the battle between twonky and WD for the configuration file supremacy, impossibility of making more then one share on external device … and many more…
So i installed Omv 1.16 that got bricked after changing the datetime and … hostname… bazinga… the three windows Ping, Tftp and Dhcp solve the problem and thanks to Fox help i dd the Omv 1.17 image… reboot … nice all was working.
I changed the hostame, enabled SMB … reboot … hello … hello … nobody on the network… Ping, Tftp, Dhcp … i’m in again and added (in bold)
/etc/network/interfaces0
auto lo
iface lo inet loopback
iface lo inet6 loopback
auto eth0
allow-hotplug eth0
iface eth0 inet dhcp
Do not work, light is white instead of green but nothing, router do not see the WDMC.
Just for clarifying, installing OMV just created a new partition ? md126 and md127 where both here in stock install ?
Where does the device boot when pushing the rest hole for more then 40s at boot ?
1 Like
I may have spent a couple of days struggling to understand that the Makefile of the Kernel always reverts to use the /linux/arch/arm/configs/sequoia_defconfig no matter how many “make config” and “make menuconfig” you do!
So if you want to build a custom Kernel just replace the **bleep** /linux/arch/arm/configs/sequoia_defconfig
This has nothing to do with cross-compilers used. I think I tried and even compiled the all lot ot them for the sake of this
Recent Kernel 3.2.68 (source supplied in Google Drive Folder, see first post) is failing to compile close to the end with the following:
make[3]: Leaving directory /home/pat/cloud/src\_linux-3.2.68-wd/\_bld' make[2]: Leaving directory /home/pat/cloud/src_linux-3.2.68-wd/kmodules/mspd-c2k/pfe/pfe_ctrl’
arm-unknown-linux-gnueabihf-strip --strip-debug -o pfe.ko ./pfe_ctrl/pfe.ko
make[1]: Leaving directory /home/pat/cloud/src\_linux-3.2.68-wd/kmodules/mspd-c2k/pfe' **make -C pws/kmodules/iscsitarget-code KERNELDIR=pwd`/_bld**
/bin/sh: 1: pws: not found
make: *** /kmodules/iscsitarget-code: File or folder does not exist. Stop.
make: *** [pfe] Error 2
Yeeeahh…
Just remove all iscsi lines from Makefile or use Makefile from previous version.
edit: nm, kernel is just copied from gpl source already released
can I use this tutorial for a WD MyCloud Mirror as well?
No. MyCloud Mirror have different firmware (Kernel in dedicated microchip).
But its possible to port. But i dont have Mirror and cant do this.
1 Like
ok, too bad but thank you for your fast reply!
