Share Folder Empty - Files still exist and accessible

[sawdustfarmer]

I thought so, what do you suggest I do next?

[JorgeB]

Post a SMART report for old disk1, or new diags with the disk connected.

[sawdustfarmer]

Attached is the SMART report after doing an extended self test on Disk1

 

tower-diagnostics-20181118-0835.zip

 

tower-smart-20181118-0835.zip

[JorgeB]

Disk passed the extended test which is good news, but it's showing previous read errors, it should be OK to do a new config with the old disk and resync parity, though if you have a spare rebuilding the current emulated first is a good option to leave you with more options.

[sawdustfarmer]

I have a spare disk that has already been precleared and I have started the rebuild.

 

It now says "Unmountable disk present: Disk 1 • WDC_WD80EFAX-68KNBN0_VAG4EY0L (sdd)" is this normal? It no longer says the failing 4TB I removed.

tower-diagnostics-20181118-1159.zip

[JorgeB]

9 hours ago, sawdustfarmer said:

It now says "Unmountable disk present

It's expected since the emulated disk was the same, don't format, when the rebuild finishes remove that disk and keep intact for now, then do a new config with the old disk and resync parity, any doubts ask.

[sawdustfarmer]

After 2 days the new 8TB disk has been rebuilt, it still says 'Unmountable: No file system' and the files that were on the failing 4TB are still missing from the array.  (they are still on the failing 4TB though)

 

What should I do next? I don't really want to do anything without clear instructions as things seem to be working better now. 

tower-diagnostics-20181120-1938.zip

[JorgeB]

-Tools -> New Config -> Retain current configuration: All -> Apply

-assign old disk1 instead of current one

-start array to resync parity

[sawdustfarmer]

Is it ok to physically take out the new 8TB for this operation? I don't have any free sata ports on my HP Gen8 so I'll need to remove it to make room for the old 4TB Disk1

 

Once I assign the old 4TB Disk1 will it try and set it up as a new drive like it did last time and erase all the files off it?

[trurl]

4 hours ago, sawdustfarmer said:

Is it ok to physically take out the new 8TB for this operation? I don't have any free sata ports on my HP Gen8 so I'll need to remove it to make room for the old 4TB Disk1

Yes

4 hours ago, sawdustfarmer said:

Once I assign the old 4TB Disk1 will it try and set it up as a new drive like it did last time and erase all the files off it?

Not if you follow those instructions exactly.

4 hours ago, johnnie.black said:

-Tools -> New Config -> Retain current configuration: All -> Apply

-assign old disk1 instead of current one

-start array to resync parity

This will resync parity with all the disks exactly as they are now. If it gives you a check box to format anything, don't check it.

 

The whole point of this exercise is to get your array going again with the files on that disk intact.

[sawdustfarmer]

Parity sync has started and all the files seem to be there except the files in the Movie share are missing again, I can see all the files through MC in /mnt/disk1, 2 and 3, the Unraid web interface sees there is 3.36 TB of files over the 3 drives but reports 0 objects: 0 directories, 0 files.  Windows 10 also sees an empty share folder

[JorgeB]

Post current diags

[sawdustfarmer]

tower-diagnostics-20181122-2036.zip

[JorgeB]

Everything looks fine to me, if you browse to /mnt/user/Movies with e.g. mc do you see the files? If not wait for the sync to finish, reboot and try again.

[sawdustfarmer]

/mnt/user/Movies is empty, its got just over 24 hours left in the rebuild so i'll reboot after the rebuild.

 

When you say try again do you mean check /mnt/user/Movies -or- do new config and resync parity?

[JorgeB]

8 hours ago, sawdustfarmer said:

check /mnt/user/Movies

This, and if it still doesn't work post new diags after trying to access it.

[sawdustfarmer]

Rebuild is complete, I checked a few files that I could not copy (through MC) off the failing 4TB onto my portable drive and they work perfectly ok now

after a reboot /mnt/user/Movies is still empty, Unraid GUI reports 3.36TB but 0 objects: 0 directories, 0 files and I can't see anything in the folder through Windows explorer

 

Strangely Plex can access and play the files in the Movies directory without an issue, except for the latest movie that was added on October 24 it can't find that file even though I can see it on /mnt/disk2/Movies.

tower-diagnostics-20181125-1550.zip

[trurl]

Your log is full of ACPI errors. See way down in this post for a possible solution:

 

https://forums.unraid.net/topic/59375-hp-proliant-workstation-unraid-information-thread/

 

Other than that I don't see anything obvious, but I suspect something is breaking your User Share(s). Do you have other User Shares that aren't working?

 

I think I have seen this sort of User Share problem with cache problems such as filesystem corruption.

 

Maybe someone else will have some more ideas.

 

*added*

 

I see you have btrfs for your array disks also. Don't know if this can be involved or not but maybe those disks might be the problem instead of cache.

[sawdustfarmer]

I tried to stop the Array so I can fix they ACPI error and its stuck and says "Array Stopping•R etry unmounting disk share(s)..." going into MC I only see /mnt/disk2 /disk3 /disks. /disk1 and /user look like they've unmounted.

 

Is it safe to shutdown? I don't want to screw up my file system again with an unclean shutdown.

tower-diagnostics-20181125-1620.zip

[trurl]

If you have console/telnet/ssh open and the command line or MC with a current working directory on a disk or share that will prevent unmounting since it will have the folder open.

[sawdustfarmer]

Looks like I had PuTTy open with MC when I tried to unmount. I've closed it and its still stuck retrying, is there something else I should be doing?

[sawdustfarmer]

I ended up clicking reboot in the GUI and it rebooted. I fixed the ACPI errors

 

I just saw in my log though

Nov 25 18:31:59 Tower kernel: BTRFS error (device md1): parent transid verify failed on 582139904 wanted 116345 found 116343
Nov 25 18:31:59 Tower kernel: BTRFS error (device md1): parent transid verify failed on 582139904 wanted 116345 found 116343

 

tower-diagnostics-20181125-1832.zip

[JorgeB]

Filesystem has generations errors, likely still mounting because the generation is very close, those type of errors can't be fixed and are usually the result of an unclean shutdown, you should backup that disk (to another one on the array or to an external disk), reformat and restore the data.

[sawdustfarmer]

Is ‘device md1’ disk1? if so that’s the failing 4tb drive I’ve been having issues with. I have a new 8tb which I can swap in and rebuild parity.

I still don’t know how to get my movies share folder back though

[JorgeB]

Just now, sawdustfarmer said:

Is ‘device md1’ disk1?

Yes.

 

Just now, sawdustfarmer said:

I have a new 8tb which I can swap in and rebuild parity

That won't work, as rebuilding from parity will not fix the filesystem issues.

 

1 minute ago, sawdustfarmer said:

I still don’t know how to get my movies share folder back though

Likely related to the filesystem issues.