RN316 does no finish booting and OS install gives error "could not extracting files) last FW

Hi all,

is there anyone who knows how to mount an USB drive and how to move the data from the full partition in there?

What about how to start SMB when the partition is full?

thanks,

---

helio wrote:

 

What about how to start SMB when the partition is full?

 

---

The question on this is how to start SMB from tech support mode.

StephenB 

apologies for the delay, reinstalling my PC and many audio video editing apps under Win 11.

I just want to say, Hope you are recovering well and presto.

I came from Windows world, I don't know anything here, how can this even happen? how is it possible that a 3rd party app is allowed to write in the sys partition?

Dont get me wrong, in principle, I would say I like linux, fedora and family. I just dont know a thing and really need to get my tons of data back; Thanks for all, thought.

Best,

---

helio wrote:

I just want to say, Hope you are recovering well and presto.

 

---

Thx. I am improving, but still have limited mobility.

However, I was able to boot my test NAS (an RN102) into tech support mode today, and mount an NTFS USB drive.

These instructions assume that you are still in tech support mode, with the data partition mounted:

Start by connecting the USB disk.

Now type

modprobe ufsd
lsblk

lsblk will give you a list of all the disks. We need to look at this list, find the USB drive The list on my test NAS is as follows:

NAME      MAJ:MIN RM   SIZE RO TYPE  MOUNTPOINT
sda         8:0    0 931.5G  0 disk
|-sda1      8:1    0     4G  0 part
| `-md0     9:0    0     4G  0 raid1 /
|-sda2      8:2    0   512M  0 part
| `-md1     9:1    0 511.4M  0 raid1
`-sda3      8:3    0   927G  0 part
  `-md127   9:127  0 926.9G  0 raid1 /data
sdb         8:16   0 931.5G  0 disk
|-sdb1      8:17   0     4G  0 part
| `-md0     9:0    0     4G  0 raid1 /
|-sdb2      8:18   0   512M  0 part
| `-md1     9:1    0 511.4M  0 raid1
`-sdb3      8:19   0   927G  0 part
  `-md127   9:127  0 926.9G  0 raid1 /data
sdc         8:32   1  14.9G  0 disk
`-sdc1      8:33   1  14.9G  0 part
mtdblock0  31:0    0   1.5M  1 disk
mtdblock1  31:1    0   512K  1 disk
mtdblock2  31:2    0     6M  1 disk
mtdblock3  31:3    0     4M  1 disk
mtdblock4  31:4    0   116M  1 disk

My test NAS is a 2-bay system with 2 hard drives (sda, sdb).  The USB drive is showing up as sdc.  Yours likely will be the last sd drive listed, just as mine is. 

Note that the other sd* disks all have a partition that is part of the data volume (sda3, sdb3 in my case end with RAID1 /data).  The USB disk doesn't say this.   

Make a note of the drive partition name - that is on the line below the sdc in my listing: sdc1.  Some USB drives have more than one partition.  If you have multiple partitions, then the NTFS partition is almost certainly the largest one. 

If you are stuck here, you can post the output of lsblk, and we can help identify (or confirm) the USB drive.

Now to mount it.  We no longer need the /mnt mount point for the OS partition, so we can get rid of that

cd //
umount /mnt

Now mount the partition on the USB drive - note in my case that is shown as sdc1 by lsblk (not sdc).

mount -t ufsd /dev/sdxx /mnt

Note you don't type sdxx - that is a placeholder.  Instead you are typing the USB partition there (in my case sdc1).

The USB drive should now be mounted as /mnt.  You can then see the root of the drive by entering

ls -als /mnt

After you get to this point, we can give you help on copying files to the drive.

---

helio wrote:

 

I came from Windows world, I don't know anything here, how can this even happen? how is it possible that a 3rd party app is allowed to write in the sys partition?

 

---

This is an embedded system, so it doesn't have all the safeguards to protect the OS from users that Windows has. Though it would have been better if the apps were running in a container (like Docker), so they would not have as much ability to do damage.  Though back when OS-6 was architected (~2013), the need to do that wasn't as clear as it is today.

StephenB I am there, does it look ok? if so ready to move files:

root@3C62350D01347://# ls -als /mnt
total 0
0 drwxr-xr-x 1 root root 0 Oct 12 19:22 System Volume Information
root@3C62350D01347://#

---

helio wrote:

does it look ok? 

root@3C62350D01347://# ls -als /mnt
total 0
0 drwxr-xr-x 1 root root 0 Oct 12 19:22 System Volume Information
root@3C62350D01347://#

---

Yes, this looks good.

---

helio wrote:

if so ready to move files:

 

---

Do you recall if you had snapshots enabled?  If so, were they available on Windows?

The linux command to copy files is the cp command.  if you want to copy everything on the NAS data volume you can enter

cd /mnt
cp -rpv /data .

rm ap

If instead you want to copy everything in a specific share, then enter

cd /mnt
cp -rpv /data/sharename .

using a real sharename of course.

Note the v option in -rpv stands for verbose.  That will list the files as they are being copied. If you don't want that listed, then you can must leave the v out.

StephenB 

This is not going to do it.

The USB drive is `-sdg1      8:97   1  57.6G  0 part

but the RN 316 has 6 disks of 4 TB, close to 14 Tb used of a 18 Tb total capacity.

I thought we were going to move the files out of the system partition, to allow the OS reinstall to take place or somehow use the newly mounted drive to allow to delete these files.

The only place I could copy the data to is on the new SAN and then reset the system and start over.

Is there anyway to delete the files from sys partition? I remember a few years back I had the same issue and the support person was able to login remotely over a SSH Technical support mode and delete the files allowing for the OS to be reinstalled.

The alternative would be to get the RN 316 to talk to the PC over netbios.

There is got to be a way.

---

helio wrote:

I thought we were going to move the files out of the system partition, to allow the OS reinstall to take place or somehow use the newly mounted drive to allow to delete these files.

---

Sorry, I thought the goal was to back up the data volume (at least files that are not replaceable), and then do a factory default.

---

helio wrote:

 

Is there anyway to delete the files from sys partition?

---

Our attempts to delete files there all failed, due to lack of free metadata (all those " No space left on device" errors).

Netgear can remotely rebuild it (destroying the OS partition and completely recreating it).  But I do not know the steps that they use.

See this article for some potential ways to fix the problem: https://marc.merlins.org/perso/btrfs/post_2014-05-04_Fixing-Btrfs-Filesystem-Full-Problems.html .  But since you are messing with the OS partition, you really need to be sure you have a backup in case it goes even more sideways and you have to factory default.

I have a process that should work using export and import as well, but I don't know if you can export in Support mode.  And if it works, you'll forever end up in FlexRAID mode, not XRAID.  Let me know if you want to try it and I'll list the details.  The first question would be whether volume_util is an available command to do the export.

All in all, making a backup and then doing a factory default is undoubtedly the cleanest solution.

---

Sandshark wrote:

The first question would be whether volume_util is an available command to do the export.

 

---

FWIW, it's a soft link to rnutil, which does run in tech support mode.  

hi all, got a 7 TB disk mounted (I think):

please, help to create and set destination target, thanks:

sdh 8:112 0 7.3T 0 disk
`-sdh1 8:113 0 7.3T 0 part
root@3C62350D01347:/# ^C
root@3C62350D01347:/# mount -t ufsd /dev/sdh1 /mnt
root@3C62350D01347:/# ls -als /mnt
total 0
0 drwxr-xr-x 1 root root 0 Oct 20 17:06 System Volume Information
root@3C62350D01347:/# cd /mnt
root@3C62350D01347:/mnt# cp -rpv /data/documents
cp: missing destination file operand after '/data/documents'

i found this on the web 

# cp -r source_directory destination_directory

launched, is still executing but it says: "operation not supported" ?

Sandshark 

What does mean "you'll forever end up in FlexRAID mode, not XRAID. " ?

I was using FlexRaid already anyways but not sure to understand. Even after a factory reset it would no longer be possible to change back to XRaid?

As far as I understand what is happening to me would happens either way. This is a extremely dangerours system that should not be trusted with any important data on any occasion, is my conclusion.

"Let me know if you want to try it and I'll list the details."

I am ready to try any non-destructive operation that could rescue my data.

Right now is running the copy document, i think but not sure it is doing anything then scan file by file . I am waiting to then plug the disk back in PC and see if copied anything.

cheers

the operation stopped at the folders starting with R, meaning appears to have copied from A to R, dont know why it stopped. already near 160 Gbs. there was about 650 Gbs total in that Documents folder.

How to restart from the starting with R Folders to the end?

I unplugged the drive and connected to PC, now copying these to the PC to see if they are functional.

Will try the same for every folder in Data, knowing will lose the Backup folder, a few TB of backup in there, not going to fit in the disk.

Still dont' know if I should be crying or laughing, still resenting Netgear for not documenting this incident type and its solution, as only a bunch of mercenaries would do.

---

helio wrote:

root@3C62350D01347:/# cd /mnt
root@3C62350D01347:/mnt# cp -rpv /data/documents

---

You might have gotten past this, but you missed a dot at the end of the cp command.

cp -rpv /data/documents .

That dot says to use the current folder you are in as the destination folder.

You could also use /mnt instead of the .

---

helio wrote:

How to restart from the starting with R Folders to the end?

---

One thing you can do is to add u to the options

cp -rpvu /data/documents .

or

cp -rpvu /data/documents /mnt

The u (update) option tells cp to only copy the files/folders if

• the file is missing from the destination (your case)
• the file on the source is newer than the file on the destination.

---

helio wrote:

Sandshark 

What does mean "you'll forever end up in FlexRAID mode, not XRAID. " ?

 

My method is to export the current volume, do a complete new set-up with a single drive (effectively a factory reset with that one drive), then import the original (degraded, since you have only 5 free slots), then export the volume created on the new drive.  Then you have room to put in the 6th drive for the original (now imported) volume to re-sync and become redundant.  But you can only do the import in FlexRAID mode.  And even after exporting one volume, you'll be unable to go to XRAID because you have "an expanded volume".

 

What this does is create a clean OS on the new boot drive, which should then be copied as RAID0 to the others once you do the import.  There are a few caveats I've not covered, so I can give you more detail if you do want to try this.  I have, BTW, experimented with this process using my "sandbox" NAS, and it worked.  But there are a lot of things that can go wrong, thus the need for backup.

 

If you just do a factory default with all drives and restore from backup, then you can certainly be in XRAID.

 

 

---

hi all,

StephenB Hope you feeling better!

I am feeling so grateful!!! Even if I might not get all the data back, I think I am getting most of it. Got 5 putty consoles up and copying data right now.

Copying most of the data to the USB 7 TB disk, from there to the new Asustore AS6510T.

This one got 2 NVM .2 disks as cache and 24 GB ram, 29 Tb , atom 4core proc.; 2 x10 Gbs and 2x 2.5 GB NICs and its Linux OS is way ahead.

Once done, I am planning to install a new fan in the RN316. I saw on the web that Netgear has inverted and remixed the standard fan connectors so that it is needed to re-order the cable pins in the fan so not to short-circuit it and burn the fan.
It might be the same for the power supply, which I am also thinking might be a good idea to replace it.

Anyone confirms this and know how to do it?

All comments welcome, thanks a lot everyone. cheers,

---

helio wrote:

StephenB Hope you feeling better!

 

I am feeling so grateful!!! Even if I might not get all the data back, I think I am getting most of it. Got 5 putty consoles up and copying data right now.

Great!  Good luck with the new NAS.

I'm definitely doing better, but it will take some weeks to fully heal.

Hi all,

StephenB I am glad you are getting better, collarbone takes long to heal, be patient.

 there was a pb with the copy, it only copied 95 Gb out of terabytes. I was looking to the 7 Tb disc to be full. I am not sure what happened.

I will try again, this time I will only start one job at time. I guess what I need is to move the data to the disk this way the next job can move the data left over.

Could you help with the syntax for the move job, please? Thanks.

BR

---

helio wrote:

 

Could you help with the syntax for the move job, please? Thanks.

 

---

Looking back at the command I posted before, I see I made a mistake: the -r should have been a -R.  So you can just use

cp -Rpuv /data /mnt

Note the -u option.  That means "copy only when the source file is newer than the destination file or when the destination file is missing".  So it won't recopy the 95 GB that's already been transfered.

Rather than moving files, I suggest using these commands to double-check that everything is copied:

find /data | wc -l
find /mnt | wc -l

The first command gives you the number of files on the data volume, the second gives the number of files on the USB drive. These can take a little while, as they are scanning the entire file system. 

You can also check total storage using

du -sh /data
du -sh /mnt

These also can take a while.  They might not match exactly, since you have two different file systems (NTFS and BTRFS). 

The -h says to give the results in "human readable" units (M for MiB, G for GiB, T for TiB)

StephenB Brilliant, thanks for sharing it. nice to be able to verify and copy only the newer data.

The pb is the data will not fit in the disk, unless I move it out to the PC to make room for the next batch, each time it is moved to the disk. The data is not being updated on the PC or the SAN.

The idea is that I move part of the near 13 Tb data to the 7 TB disk, move what's in the disk to the PC, to make room for the rest of the data left in the SAN, as the 13 Tb data won't fit in the 7 TB disk. does it make sense?

Thanks and BR,

---

helio wrote:

 

The idea is that I move part of the near 13 Tb data to the 7 TB disk, move what's in the disk to the PC, to make room for the rest of the data left in the SAN, as the 13 Tb data won't fit in the 7 TB disk. does it make sense?

 

---

Ok.  But I still am hesitant to move data, because it deletes it from the source.  If something goes wrong, you end up with needless data loss.

Can you post the output of 

du -h -d 1 /data/

 If the largest share will fit on the USB, then we could copy one share at a time.

As long as no single share or directory has more than the 7TB so you can copy a sub-section of files each time, that "bucket brigade" method should work well.  If you have one that exceeds that, then copying alphabetically might help select a small enough group.  Just remember that in Linux, upper/lower case matters.

Back in ancient times, we called it "sneakernet":  carrying disks (at that time, floppy ones) between computers to move files (wearing your sneakers). 

StephenB , this is great, I wonder what is causing the copy of Documents stop in the middle when it encounter an old folder containing a copy of a USB drive..

How to copy things alfphabetically?

I am currently using folder names for copying, eg., 

# cp -rpv /data/Software .

root@3C62350D01347:/# du -h -d 1 /data/
0 /data/home
18G /data/.apps
0 /data/.vault
1.9M /data/._share
6.2G /data/Pictures
71G /data/Videos
270G /data/Music
531G /data/Documents
0 /data/.timemachine
482G /data/Freaks
20G /data/Documentation
1.9G /data/Podcasts
254M /data/eBooks
36M /data/Transfer
488G /data/Audio
16G /data/HdAProdz
1.4T /data/Software
130G /data/OneDrive
8.4T /data/BackUp
0 /data/AmazonCloudDrive
8.0K /data/DropBox
16M /data/GDrive
0 /data/.purge
1.4G /data/uFreak
407M /data/ODB
14G /data/DZI
453G /data/VSTIData
0 /data/Video
13T /data/

Thanks,

---

helio wrote:

 

root@3C62350D01347:/# du -h -d 1 /data/
...
8.4T /data/BackUp
...

 

---

Ok.  So everything except BackUp will fit on the 7 TB drive - which is probably 8 TB (~7 TiB).

Unfortunately, there is no simple way to exclude a folder with the cp command - but we can easily switch to rsync.

rsync -avzh --exclude '/data/BackUp' /data/ /mnt/

Note you can enter this after your current copying is done.  Rsync is incremental, it will skip the folders you've already done.

---

helio wrote:

I wonder what is causing the copy of Documents stop in the middle when it encounter an old folder containing a copy of a USB drive..

 

---

I don't know, but it is worth looking again after using the rsync command - that might manage to get past this problem.