On 09/11/2015 11:56 PM, Mike Miller wrote:
>
> [169448.936046] ata3.00: error: { UNC }
> [169449.048405] ata3.00: configured for UDMA/133
> [169449.048439] sd 2:0:0:0: [sda] FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
> [169449.048442] sd 2:0:0:0: [sda] Sense Key : Medium Error [current] [descriptor]
> [169449.048445] sd 2:0:0:0: [sda] Add. Sense: Unrecovered read error - auto reallocate 
> failed
> [169449.048447] sd 2:0:0:0: [sda] CDB: [169449.048449] Read(10): 28 00 73 c5 1d 58 00 00 
> f8 00
> [169449.048456] blk_update_request: I/O error, dev sda, sector 1942298138
> [169449.048470] ata3: EH complete
> [169449.048603] JBD2: Failed to read block at offset 8899
> [169449.048900] JBD2: recovery failed
> [169449.048903] EXT4-fs (sda1): error loading journal
>

I just wrote you a different response last night, where I noted that ext4 is the only one 
that cost me data... and here again, a single IO error on your drive, and ext4 has puked.

It used to be that you could mount ext4 as ext3 - which would cause it to ignore the 
journal - but I think they have since changed the format of ext4, and that is no longer 
possible (unless you created this ext4 a really long time ago, when they were still 
compatible, by default)

As far as trying anything else for recovery - the first step is to get whatever is on this 
drive onto good hardware - as any further operations may lead to more damaged blocks, 
making the problem even worse.

Use a tool like one of these, to copy the entire disk, bit for bit onto that nice, new 
empty drive that you have:

http://askubuntu.com/questions/211578/whats-the-difference-between-ddrescue-gddrescue-and-dd-rescue


Its possible, that after copying the entire disk, bit for bit - that you might be able to 
mount it from the new drive - because the rescue tools do a much better job of isolating 
bad bits (as opposed to failing an entire sector read), and perhaps the remaining errors 
will be small enough to let ext4 complete the mount.

If not, start going down one of these recovery paths:
https://linuxexpresso.wordpress.com/2010/03/31/repair-a-broken-ext4-superblock-in-ubuntu/

But ONLY on the new drive - after you copied the data.  The failing drive could completely 
disappear on you at any time.

After you get done doing all of this, and hopefully get back the data you want - then you 
can go down the road of reformatting that new disk the way you want it in the end.

 From your other message:

> I think parted is pretty easy to use, once you know what to do.  I guess I need to use 
> parted -l to find the /dev/sd letter (I'll use /dev/sdx here), then something like this 
> will create the ext4 partition:
>
> sudo parted /dev/sdx mklabel gpt
> sudo parted /dev/sdx mkpart primary ext4 0 5TiB
>
> I think it then needs to be formatted.  Would this command do it?
>
> mkfs.ext4 /dev/sdx1

This sounds about right - but all of this will be much easier if you just use gparted.

Also, I'd still recommend doing an encryption step.  So, after you make the primary 
partition, with truecrypt, anyway, you encrypt the entire primary partition.  Then, after 
you "mount" the encrypted partition with truecrypt, you do the file system create steps 
inside of that encrypted container.

I'm sure other encryption tools are similar, I'm just not familiar with them.

You can find the last good version of truecrypt here: 
https://www.grc.com/misc/truecrypt/truecrypt.htm including the user guide: 
https://www.grc.com/misc/truecrypt/TrueCrypt%20User%20Guide.pdf