NAS326 problem
I need help and advice. I have NAS326 with latest firmware. Some days ago error appeared:
Then automatically web interface goes to this:
As you can see, there is no option "repair"
There are also these status pictures:
and this:
What should I do, how can I save all data stored on this NAS ?
Any advice please ?
Thanks
#NAS_June_2020
All Replies
-
How long has the disk been using? Perhaps it's about to die.0 -
@Scofield - As you could see there is no "repair option".
I am using this NAS about 5 years without problems.
I am seeking any advice, I can use putty but I do not know special linux commands how to see what is going on, how can I see mounted disks and how to "check" disks on NAS326 via putty.
Simply, any advice how to save data.
Thanks,
0 -
Can you login over ssh, and post the output ofcat /proc/mdstatcat /proc/partitionscat /proc/mountssumdadm --examine /dev/sd[ab]30
-
Did I lost all my data ?
Yes, I'm afraid you did.
/proc/partitions shows all your block devices (disks, partitions, ...), and it shows that both your harddisks, sda and sdb have 3 partitions, as it should.
/proc/mdstat shows the status of the 'multi device' devices (raid arrays). There should be 3 of them, there are only two. (system and swap). Both remaining arrays show their partition on sda as faulty. According to 'mdadm sda3' there is no raid header on sda3. So somehow sda got corrupted on 3 raidheaders on 3 partitions. So I'm afraid that disk is dead, although the partition table is still readable.
According to 'mdadm sdb3' your data partition was a linear array, consisting of 2 partitions. Which means the partitions are glued after each other, with no redundancy. The first half is gone, so now only the second half of the filesystem remains. I am not aware of a way to repair/mount that half.
There are 2 possibilities left. You can use low-level recovery. PhotoRec and similar software can recover files without help of the filesystem, by scanning the raw partition/disk for recognizable files. Problem is that the files have to have a clear header, and so it can recover most media files and office files, but flat text is unrecoverable. Further it can only recover files, not metadata (which is part of the filesystem), and so filenames, directorynames and timestamps are lost.
The other possibility is recreate the data array, and hope the filesystem itself is not beyond repair. But as all raidheaders on sda are damaged, which represents a span of 4GB, I think the odds are bad. Yet it won't hurt to try.
If your data is valuable, go to a specialized data recovery business. That will cost hundreds of euros. If it's not that valuable I'd first try PhotoRec (which is available for a lot of operating systems, and as it doesn't use the filesystem, the OS doesn't need to support it), and when you have recovered everything you can, try to recreate the array (and yes, I can help) and repair the filesystem.
The reason for this sequence is that PhotoRec leaves the data as it found it, while trying to repair a damaged filesystem can add extra damage.
0 -
@Mijzelf
Well,
i have been 20 years ago a linux oriented...and there was and fsck command that was used to check, and optionally repair, one or more filesystems.
Is this an option that could help ?
And I have one question - related to your comment:
The other possibility is recreate the data array, and hope the filesystem itself is not beyond repair. But as all raidheaders on sda are damaged, which represents a span of 4GB, I think the odds are bad. Yet it won't hurt to try.
Please, could you be more specific how to do this ?
Thanks0 -
fsck is a tool wich you run on a block device (something containing a filesystem). The block device is in this case virtual, and absent, as it is the raid array containing 2 partitions which cannot be assembled.(Which I think is the reason there is no repair button. The firmware can't find the volume)Please, could you be more specific how to do this ?The command ismdadm --create --assume-clean --level=linear --raid-devices=2 --metadata=1.2 /dev/md2 /dev/sda3 /dev/sdb3(That is a single line). mdadm will complain that sdb3 already is part of an array. After creating there will be an array /dev/md2 which might be mountable. If not, you can run e2fsck on it.
0
Categories
- All Categories
- 415 Beta Program
- 2.4K Nebula
- 146 Nebula Ideas
- 96 Nebula Status and Incidents
- 5.7K Security
- 262 USG FLEX H Series
- 271 Security Ideas
- 1.4K Switch
- 74 Switch Ideas
- 1.1K Wireless
- 40 Wireless Ideas
- 6.4K Consumer Product
- 249 Service & License
- 387 News and Release
- 84 Security Advisories
- 29 Education Center
- 10 [Campaign] Zyxel Network Detective
- 3.5K FAQ
- 34 Documents
- 34 Nebula Monthly Express
- 85 About Community
- 73 Security Highlight