Lacie 5Big2 Failed After Firmware Update

[fvdw]

I am at work right now, so can not be online all the time
My knowledge of raid is minimal but problem of not having mdadm in the standalone kernel is no a real problem we can upload that via tftp.
I will post it this evening
ps even if you have raid you still should be able to mount /dev/sda2 and look to the content
did you try ? ( mounting will not affect data on it)

[Mijzelf]

even if you have raid you still should be able to mount /dev/sda2 and look to the content

Don't think so. That is (sometimes) true for a raid1 array, where both disks are equal, and so both contain a valid filesystem. But on a raid5 system the filesystem is striped over several disks, and so the filesystem cannot be mounted without assembling the array.

BTW, you'll also need kernel support for md devices *and* for the specific raid level.

[unlucky1]

Ugh! I feel like I'm so close yet so far! I just need to get close enough to get my data safely backed up to another drive. I think your tool is my last hope! Maybe on the lacie there's a byte I can tweak to get to the we dashboard again so I can remount my files?

Thanks for both your help so far! fvdw, you don't happen to work at FoxIT, do you? It just seems like most NL IT people I run into happen to work for them ;)

[fvdw]

no I don't work at FoxIT, I am not a professional IT specialist only doing this as a hobby

The standalone kernel supports md devices and all raid levels (you should be able to see that in dmesg output)

Code: Select all

dmesg

ps mdadm might need libc, the standalone kernel only contains static versions of binaries, otherwise it becomes to big.
But the 5big2 has 512 MB ram should should be able to use the libc libraries as well as RAM is big enough to store them. This evening i will have a look

Maybe on the lacie there's a byte I can tweak to get to the we dashboard again so I can remount my files?

maybe yes, the Lacie firmware runs from partition 7 and 8, those should be possible to mount and adapt.
However I don't use the crappy firmware of lacie so I don't know what should be checked there and/or changed to let it do its thing.
Maybe Mijzelf can help here.

[unlucky1]

Thanks again for your help!

[Mijzelf]

Maybe on the lacie there's a byte I can tweak to get to the we dashboard again so I can remount my files?

maybe yes, the Lacie firmware runs from partition 7 and 8, those should be possible to mount and adapt.
However I don't use the crappy firmware of lacie so I don't know what should bechecked there and/or changed to let it do its thing.
Maybe Mijzelf can help here.

I'm sorry. I know that on a Networkspace2 Unicorn is not happy if it fails to mount the data partition, causing the dashboard to fail. The remedy is to make it mountable, which in most cases means reformatting it. In this case that's *not* a solution.

BTW, partitions 8 (and 9) are also in a raid array, a 5 disk raid1 array (yeah, redundancy is it). I'm not sure about 7, but that one is not responsible for the problem. So you could mount a single partition and change something, but I think the changes will just be reverted by re-syncing the array.

[fvdw]

I compiled mdadm static so it won't need glibc

Two commands can not be used:
'getgrnam' and 'getpwnam'
in statically linked applications these will require at runtime the shared libraries from the glibc version used for linking
I used glic 2.3.6
But maybe you don't need those commands
Attached the binary in a zip archive, extract and put the file it in the tfp sub folder of the fvdw-sl console
You can upload it to the standalone kernel using in telnet the command (assuming you started the tftp server and have put the file in the tftp directory of the fvdw-sl console) using this command

Code: Select all

tftp -l /sbin/mdadm -r mdadm -g ip-pc

(replace ip-pc in this command by actual ip of the pc running fvdw-sl console)
After it is uploaded make it executable

Code: Select all

chmod 755 /sbin/mdadm

mdadm is now available here:
http://plugout.net/viewtopic.php?f=7&t=1271&p=18801#p18801

[unlucky1]

Hi,

Thanks so much for getting me that program. It works! But a new problem:

mdadm: cannot open device /dev/sdb2: No such device or address
mdadm: /dev/sdb2 has no superblock - assembly aborted

dmesg:

[ 6002.519019] sdc1: rw=0, want=2052, limit=2048
[ 6002.525934] UDF-fs: No partition found (1)
[ 6767.383338] UDF-fs: No partition found (1)

What's a superblock? Next steps? Any ideas? That doesn't look good!

[unlucky1]

Based on what I found here:

https://www.linuxquestions.org/question ... it-932916/

Here's some more info in case it helps:

root@(none):/ # cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
unused devices: <none>
root@(none):/ # mdadm -D /dev/md0
mdadm: md device /dev/md0 does not appear to be active.

Is it worth trying what they suggest or do you have any better advice? I really can't toast this data!

[Jocko]

Hi unlucky1,

You have some issue with your partitions...

So you need to repair them : Check if you have the command e2fsck.

Code: Select all

which e2fsck

If yes, try this

Code: Select all

e2fsck -y /dev/sdb2

if this command warns again about the superblock try this

Code: Select all

e2fsck -f -b 8193 /dev/sdb2

if not succeed replace 8193 by a value 1+nx8192 (where n=1,2,3,....)