[Solved] How to mount a drive that unraid won't mount?


Recommended Posts

See my other thread here:  http://lime-technology.com/forum/index.php?topic=13084.0

 

I've run reiserfsck and when I do I see the contents of the drive flying by.  It seems the directory structure is mostly intact.  I did get some references to Lost+Found.  However, Unraid still shows these drives as unformatted and won't mount them.  What can I do to get Unraid to bring them back in to the array, or failing that, how can I mount the drives to save the data off before formatting them through Unraid?

Link to comment

Anyone?

Does the reiserfsck finish? 

 

If you type reiserfsck --check /dev/sdX

does it pass, or does it still find errors?

 

You can attempt to mount the drive by typing

mkdir "/tmp/mountpoint"

(where mountpoint is a directory name of your choice.)

Then

mount -t reiserfs /dev/sdX1 "/tmp/mountpoint"

 

Note, it is the first partition we are mounting, so you must append the "1" to the base device name.  If the device is /dev/sdg, your mount command would be

mount -t reiserfs /dev/sdg1 "/tmp/mountpoint"

 

Link to comment

Joe, yes the reiserfsck finishes with no errors.  I copied the output of that into the other thread.

 

I tried what you suggested and got:

 

root@unraid:~# mkdir /tmp/sdc1
root@unraid:~# mount -t reiserfs /dev/sdc1 "/tmp/sdc1"
mount: /dev/sdc1 already mounted or /tmp/sdc1 busy

 

While the array was up. 

 

I shut down the array and tried again and it mounted!  I'm able to search through it at the command line now.  However, I haven't noodled how to copy the files off.  Or can I copy them all to one of my less occupied drives and then let Unraid format the "unformatted" drives?  Or does this mean I might be able to get Unraid to see them and allow them back into the array without formatting?

 

I tried starting the array again but it wouldn't let me.  Had to reboot to clear out the tmp mount.  Probably could have done a umount but oh well.

Link to comment

So, is there a way to get Unraid to mount the drives or do I have to put them in a different machine to copy the files off?

If you can mount the disk, unraid should be able to as well.  The only way it would not be able to is if you have an add-on that is using the mount-point BEFORE unRAID attempts to use it. It's mount would then fail.  Are you running any add-ons?

 

Joe L.

Link to comment

Only unmenu.  Would that be doing it?

unMENU does not mount drives automatically, that is not the issue.  (Jut make sure you have the most recent version of unMENU. You can press the "Check for Updates" button on the user-scripts page and then install them if there are any.)

 

If I had known you had it installed it would have made your task of mounting the disk easier.  On the "disk management" page there is a button to mount the external drive.  Then, once mounted, there is a button to share it on the LAN.  At that point you can copy anything you like to the other disks in your array.

 

Just don't press the button in unMENU to create a reiserfs file-system on the disk, as it already has one.  Asking it to create one is exactly like re-formatting it. (and that would delete the existing contents, so don't do that if you want to move the files off the disk)

 

Link to comment

Joe, thanks!  I updated to the latest version of Unmenu.  I was on 1.2 I think.  Now with 1.3 and I'm not seeing the "mount external drive" button on the disk management page.

Does it show you a listing of the drives not assigned to the array?  (probably it thinks the drive is assigned to the array, and therefore not showing it as a un-protected drive)

 

What does your unRAID page show?  Is your "problem" drive assigned to a slot in the array?

Link to comment

Yep, it sure is.  I took the array offline and removed those two disks but then I wasn't able to start the array due to too many disks being missing.  I tried it with just one and same result.

 

If I have the array offline and mount the disk, how can I get to it once it's mounted to copy the files off?  I might be able to scrounge up enough space outside of Unraid to move everything off if that's possible.

Link to comment

I figured out how to do that.  Press the share button.  Duh.

 

Now, is it possible to do the same with a good drive and copy between sdc1 to sdg1 shares?  Just as an example?  Or will that screw up the good drive and the parity?

It would screw up the parity if the array is not started.

 

What I would do at this point, assuming you can get to all your other drives is to set a new disk configuration and ALL are either green or blue is set a new disk configuration.  In fact, I'd leave the problem disk assigned to the array.  (re-assign it if you un-assigned it)

This is if there are no other drives that are failing (red) or marked as un-formatted.

 

To set a new disk configuration you would stop the array and then type:

initconfig

and then respond with

Yes

(capital "Y", lower case "es")

 

It will immediately invalidate parity, but it will set a new disk configuration with all the assigned and working drives.  If you refresh the unRAID management page, all the drives should show as "blue"  It will say something like it will bring the array on-line and calculate parity if you click start.

 

When you then start the array parity will be calculated on the entire array.  Your data will stay as it is.  You should NOT see any un-formatted drives, but if you do, DO NOT DO NOT format them...  Did I say DO NOT press the button marked as "format"?? 

 

Your entire array might just come back, with all your data in place. 

 

Joe L.

 

 

Link to comment

Joe, before I do this let me be clear that I have two drives that are "yellow".  The rest are green.  Both of the yellow drives can be accessed via the method you've explained to me.  The remaining drives, when assigned correctly, will allow the array to go online and be used.  So, with that in mind, do I still want to do the initconfig?

 

Thanks again for all your help!

Link to comment

Joe, before I do this let me be clear that I have two drives that are "yellow".  The rest are green.  Both of the yellow drives can be accessed via the method you've explained to me.  The remaining drives, when assigned correctly, will allow the array to go online and be used.  So, with that in mind, do I still want to do the initconfig?

 

Thanks again for all your help!

I think so. 

 

What do you mean by "The remaining drives, when assigned correctly, will allow the array to go online and be used. " ??

 

To play it safe, make a copy of your "config" directory and put it somewhere else on the flash drive.  It can be named old-config.  That way, if you type initconfig, and it does not act as expected, you can re-name the old-config to "config" and be back to where you are now.

 

Perhaps you can post a screen shot, or a current syslog, or both. 

Link to comment

Joe, you're a life saver!  Backed up config and did initconfig.  All the drives went blue.  I simply mapped them back to their original locations and started the array.  It's running a parity-sync right now but all of the drives are green.  Hopefully by tomorrow I'll know if everything's happy but it's looking much better than before.

 

Thanks VERY much!

Link to comment
  • 8 years later...

Have a similar problem, but I DO NOT want to add the drive to the unRaid array - it WAS a member of that array, but developed so many bad (reallocated) sectors that it was replaced and "rebuilt".  Problem is the "rebuild" completed with 163 (unidentifed) errors, including an 'unrecovered' CRC error between the parity drive and drive#6. (It was drive#3 that was replaced in the array.)  Therefore, the contents of the "new" drive are in question and the contents of the "old" drive are intact...

Ideally, I wish to mount the "old" drive on either the same machine as an unAssigned device WITH the Array running, or on another machine and using a SMB share, compare the contents of the 'old' drive vs. the rebuilt 'new' drive; however, THIS remains to be accomplished...

As Bitbass encounted, the "old" drive will not mount WITH the array running. Same here, EXCEPT, I could bring-up the array subsequently. BUT, the "new" drive was missing from the configuration as "unmountable" because of "no file system".  That condition was recoverered from AS SOON AS the "old" drive was not mounted along with the Array running - THERE IS AN APPARENT CONFLICT between the two drives; the "old" #3 from the array and the "new" #3, SO

The problem remains with having access to the "old" drive and the "new" drive AT THE SAME TIME, in order to inspect the contents.

Additionally, as I have another unRaid machine, I attempted to mount it there in unAssigned devices and could do so with its array running, BUT when the "mount point" was accessed, it showed NO CONTENTS, when the contents had been verified when it was mounted on the first machine.

On top to that, I installed "Linux file system for Windows" on a Windows machine - SAME ISSUE - will mount, but contents of the mount point folder are "empty".

Now, I really am stuck on how to access the data on that drive.  Any ideas, gentlemen?

(If the conflict between the two drives could be resolved, that would probably provide the quickest means of accomplishing the task. What needs to be 'altered' on the "old" drive in this case?)

 

Link to comment

Cool, Johnnie.Black.  Thanks for the info; that would explain the conflict.  Would have to do some research about 'wiping' it (if the drive would then be usable) or replacing it otherwise.

In the meantime, I got the drive to work on the 'secondary' unRaid system v6.5.3 where its operation wasn't "conflicted", but certainly "flakey".

Noticed that even though the drive was "shared", the mount point WAS NOT!  Sometimes (not always) it had to be set to "share" individually and EVEN THEN it did not always appear across the network.  Once it did, I was able to us a SMB share to access it from its system of origin and compare its contents to that of the 'rebuilt' drive that replaced it.  BIG JOB! (Unless someone knows of a better, faster way)  I'm using Krusader to generate checksum file(s) and then doing a binary compare on the checksum files between the two disks.  It works, but slow going - felt it had to be done before I 'repurpose' the drive because of the 163 (undefined) errors from the rebuild... that's all that was presented, save for the CRC error, in which case I know the bit(s) weren't read properly, so some file(s) got 'hosed' in the rebuild operation.  Not that I don't trust computers.  I trust them to be a continual source of "entertainment" in which I don't believe it until I see it...

Including operation(s) in unRaid and associated apps.  BTW on the machine to which the "old" drive is now 'mated', (binhex-) Krusader v2.7.1 STILL DOES NOT SHOW THE CONTENTS, yet they can be seen in telnet, by the SMB share, and on the "console" (Windows) machine used to control the two (headless) unRaid servers...  see what I mean about not believing it until I see it?

At any rate, thanks for the info.  I like to spell-out what I find for the benefit of others facing similar situations; if I can 'find' such problems, I'm sure someone else has probably run across them too.

 

Edited by CaryV
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.