I am pretty fastidious about moving files around but a week or two ago i was saving a few .m4v files and .. then .. i couldn't find them.  I assumed I screwed up but then yesterday I looked for several folders of scanned photos and they were just gone.  This one of course disturbed me.  Again I assumed I made a big mistake and restored a CrashPlan backup of the missing folders.  Today, however, my iTunes library is showing several broken links (i.e, missing .m4v files) so now I am pretty certain something is amiss.

 

I ran a parity check a few days ago (it had been a while; this was BEFORE the missing photos but after the missing m4v's):  no errors.  I also ran reiserfsck on all six of my drives:  the first five came back without corruption; however, the 6th disk repeatedly told me "device is busy".  I quit every device I could think of that might be connected and then rebooted, but it continued to come back with that error (as it does now).

 

Per the suggestions in the Forum, I ran the Fix Common Problems plug-in.  It came back with a Docker error (i had a container cfg in a user share not a disk share).  also some warnings regarded to disk shares (over-specifying include/exclude disks, ie. include disks 1, 3, 5; thus, therefore, exclude 2, 4, 6.  logic demands the exclusions, but i had specified them anyhow).  now fixed.  I do have a couple of drives running a bit hot (46C) and I am at 95% capacity on one (and near 90% on another two).  a SMART report is showing "FAILING NOW" on an End-to-End error on disk 2.  At any rate, I re-ran Fix Common Problems until nothing came back.  I still can not unmount disk 6.  I downloaded the Diagnostic Report (attached).

 

I have now run a Power Down.  Upon restart I CAN find the previously missing iTunes movies; however, my free disk space on my mounted shares is about 150GB larger than it had been.  And disk6 is STILL not unmountable.  So I am concerned and would very much appreciate any help.

 

One other thing I will submit is that I did connect to my shares once as a (Mac) root user via AFP (I have done this several times in the past and it has not been a problem) and deleted some stubborn files that would not delete via any sudo rm commands.  I did this after the first missing files (first paragraph) but BEFORE the observation of the missing pictures.

 

Thank you.

 

 

I am on version 6.1.9. 

 

System info:

 

M/B: ASUSTeK Computer INC. - M4A79XTD EVO

CPU: AMD Athlon II X2 250 @ 3000

HVM: Disabled

IOMMU: Disabled

Cache: 256 kB, 2048 kB

Memory: 4096 MB (max. installable capacity 16 GB)

Network: eth0: 1000Mb/s - Full Duplex

tower-diagnostics-20160525-1326.zip

How did you move the files? You should never mix disks and user shares when copying or moving files.

 

In other words, move/copy from diskA to diskB is OK, move/copy from shareA to shareB is OK, move/copy from diskA to shareB or from shareA to diskB is not OK and can result in data loss.

Thanks trurl.  The only change I made in that regard is in my Plex (Docker) container volume: when I created it I put the /config file in a user share (as I was made aware of when I ran the Fix Common Problems plug-in).  Per the FCP instructions I changed the /config host path to a disk share.  I don't know what happens to the orphaned /config file on the user share?

 

Other than that, I don't recall moving any files from a user share to a disk share or vice versa.

 

Thank you.

I never know how concerned you should be about End-to-End errors, because they are so uncommon.

 

Basically the firmware on your HardDrive likes to make sure things are working properly so it compares the checksum of the data written to the physical media with the checksum of the data in the HardDrives cache, if it doesn't match that's an End to End error. IMO I don't know that I would trust a disk that had an end to end error.

 

Additional info... not sure if this will work, but you could try

lsof | grep '/dev/sda1'

where sda1 is the device number of drive6 to see what process is hanging disk6 then kill that process.

 

You could also try a force unmount (but do this as a last resort because it won't sync the filesystem first.

Thanks trurl.  The only change I made in that regard is in my Plex (Docker) container volume: when I created it I put the /config file in a user share (as I was made aware of when I ran the Fix Common Problems plug-in).  Per the FCP instructions I changed the /config host path to a disk share.  I don't know what happens to the orphaned /config file on the user share?

 

Other than that, I don't recall moving any files from a user share to a disk share or vice versa.

 

Thank you.

Nothing "happens" to the /config on the user share. It is still there unless you went to each of your disks and deleted it. And the /mnt/diskX/config is part of that user share. A user share is just the aggregate across all disks of all top level folders named for the user share. So /mnt/disk1/config and /mnt/disk2/config and indeed /mnt/cache/config are all part of the config user share at /mnt/user/config. If you originally had it as /mnt/user/config then its files are on one or more of your disks in the config folder.

Here is an update.  Having had no luck using lsof | grep to find active processes I did a lazy unmount and proceeded to do a disk check.  It found no corruptions.  I then re-ran a parity check (0 errors).  When I used an AppleScript to check for broken links in my iTunes library, within a minute it came back with over 100 missing movies.  I terminated the script when I realized that it wasn't merely looking for movies without attached files, but deleting the entries from iTunes.  So it is likely that I have a LOT of missing movies (I haven't even checked music, nevermind photos, backed up docs, etc.).  I do believe I need to replace the disk with End to End errors, but shouldn't the parity check be finding errors?  I don't know what to do here:  it's hard to determine the scope of the problem, but I am clearly missing files, perhaps thousands or more.

update 2:  at least 400 video files missing and certainly (many) more...could i have accidentally deleted a few of these?  sure?  all of them?  no way.

How are you determining they are deleted instead of just moved somewhere else?

Fair question    Some of these are 720p mkv's:  there isn't a lot of room to put them! (my home drive is a 256GB SSD, which is usually, and not optimally, running at >90% capacity, and they certainly aren't there).  I have looked in every unRAID share that I mount on an even semi-regular basis (about a half-dozen) and I don't see them.  Of course I have looked on external drives, as well. I will say, though, that the free space figure (1.14GB) is admittedly inconsistent with the loss of hundreds of GB of files.  But if there are drive issues, this may not necessarily disprove that they aren't gone, is that right?

 

Should I be searching at the command line instead of GUI (Mac) for missing files at the drive level (i.e., execute finds on each of my six drives)?  Is this even doable?

 

Thanks.

 

You could do find from the linux command line, but I would think you could get your mac to find also. Maybe they are in some subfolder somewhere.

I have spent a few hours searching for files via GUI and command line.  In addition to over 100 movies, I am missing many more tv episodes; in fact, entire seasons of them.  Strangely and perhaps tellingly there are seasons in which I have lost all but, say, three episodes: if folders were being moved around or deleted why would 3 of 22 episodes remain?  I am trying to think of when this all might have occurred.  I will say that it does seem to coincide with my installation of the Plex docker template (it was the first time I had experimented with any docker apps).  I did NOT map any TV files to it, however.

 

Any other suggestions?  Anything in the Diagnostic Report? 

 

Thanks again. 

Depending on how you set up the shares, it is entirely possible for some episodes of the same TV series to be on a different disk.

After another power down and restart most, hopefully all, of the missing files seem to have reappeared.  I think i have confirmed gundamguy's concerns about End to End errors.

 

I am going to replace this wonky 2TB drive.  As long as I am doing so, I had might as well expand capacity.  However, my parity drive is 3TB so that is the largest drive I can replace the faulty drive with.  What are your thoughts on first replacing the parity with a 4TB drive and then replacing the faulty 2GB drive with another 4TB drive (sequentially, of course)?  Is this risky, considering the erratic behavior I am seeing?

 

Thanks yet again.

After another power down and restart most, hopefully all, of the missing files seem to have reappeared.  I think i have confirmed gundamguy's concerns about End to End errors.

 

I am going to replace this wonky 2TB drive.  As long as I am doing so, I had might as well expand capacity.  However, my parity drive is 3TB so that is the largest drive I can replace the faulty drive with.  What are your thoughts on first replacing the parity with a 4TB drive and then replacing the faulty 2GB drive with another 4TB drive (sequentially, of course)?  Is this risky, considering the erratic behavior I am seeing?

 

Thanks yet again.

 

Well I'm not 100% sure the behavior you described is what I would have expected. I don't have another answer however.

 

You might want to consider the "The parity swap procedure"

 

I've not done this personally so you might want more feedback on if this is the right path to go down and if anyone has any insight or gotcha's to avoid.

 

Edit: fixed URL