I will start off by saying the Noob posts make me realize I am about 47 levels below Noob. So feel free to tell me to go jump in a lake. Or point me toward a more appropriate forum.

I have a 2tb WD green in an external enclosure, SATA on the inside, USB2 and esata on the outside. It was a data drive in a tower for a while, formatted fat32 for large (video) files. It went portable when I got a job out of the country, and played nice with that for a few months.

Eventually, being plugged in and powered on and unplugged and switched off or ejected or having the cord pulled by me/wife/kids/friends, it will no longer reliably mount in linux, mac, or windows.

I got in once, and it showed 2tb empty. I started a test (sector by sector read? Scrape? I'll try to remember what I did exactly, but definitely read only) and it suggested it would take 57 hours. I remember the intention was for it to look for file start and end flags, and it would be able to give me those files in some format that I would then go figure out what exactly they were and get them named again.

As I didn't have 57 hours, and as it wasn't finding anything in the two hours I let it run, I cancelled the program and shut off the drive. I was told a better approach would be to get a second drive, as big or bigger, put them both in a tower rather than in external enclosures, and copy bit for bit the old to the new.

Last possible monkey in the works is that I'm in Casablanca, Morocco, where desktop/server parts are hard to get. I have fairly regular friends and coworkers back and forth every month, so I can get parts if necessary, but there isn't a Frys down the road or a Newegg or Amazon. The only desktop I've been able to put together is an old AMD socket 939, 1.5gb ram, 10/100 network, USB2, no esata. New 2tb drive is on the way, I expect to have it in about 3 weeks.

I have linux, windows, and mac laptops, so if there are any scans you recommend I do with the old drive in the enclosure under a specific OS, I can do that. I have a copy of Win XP I can put on the desktop if that OS is preferred, but I'm also reasonably comfortable in Linux (Mint is my primary OS on my laptop, Puppy on the desktop now, Ubuntu for the last few years).

I do know enough not to keep poking it while I don't know what I'm doing.

The data is all movies and music, all replaceable, so this is definitely an opportunity to learn without risking critical data.

So, on to the questions.

Should I try to do any more scans with the drive in the enclosure? I have esata on the Windows laptop, but it seems unnecessary to do this over USB2 when I have Sata in the desktop. SMART is the only kind of test I know could be informative, and I don't know if its possible the mobo I have might not support SMART? Told you I'm not even a noob.

What other info do I need to provide in order for you guys to help me?

What programs should I install and get familiar with while I'm waiting on the new 2tb drive to come in?

Is my hardware totally insufficient for this operation, and I need to build a new machine when I'm back in the US in November?

As I said at the beginning, tell me if I'm not even playing the same sport you guys are. Just trying to learn.

Given that you are happy to test this and data loss is not a "customer" issue:


Using what software?
The 57hrs is an average based upon the size of the drive and the current progress, it may alter significantly as the test time progresses.

File Carving: (is this what you meant by scrape?)
Not much good with large files unless you regularly defragged the drive.

please say that you just used slang there and the drive isn't making a scraping noise.

possible solutions just at test stage.
try hiren or ubcd and run MHDD or Victoria

Use native cables if possible, ie SATA not USB.
Run from CD with NO other drives attached.
Better safe than sorry.

Instructions here: http://real-world-systems.com/docs/MHDD_en_manual.html

Run it briefly and get a feel looking at the blocks.
If it starts filling red, come back out straight away.

Results:
Example: http://digitalferret.co.uk/hg1027.jpg

If its running in the good area then you are in luck: progress to next stages.
If not .... well, lets see first, get status codes.

Kern

_________________
Когда хочется кушать – съешь всё.
Голод не тётка!

Thanks for the reply.

Regarding the test, I don't remember off hand, it was about 6 weeks ago, I know it was a terminal based program in Ubuntu.

Regarding the word Scrape, I've been reading too much about media servers lately, I just meant "Look at all the data and tell me about it," file carving I believe is what I meant.

Do you agree with the bit for bit copy to known good drive, and try to recover data from the new drive? I believe the intent of whoever told me that was to avoid messing with the original data, the standard "edit a copy" approach.

If a copy would be advisable, I'm looking at Clonezilla or terminal, dd if of. Advice?

test first, then if you go to dd or a derivative you will know if/why its choking.


Maybe use hexdump or similar too, if the drive isn't stalling to see what data it's pulling out.

You probably know anyway but don't mount the drive just direct the command at /dev/sda or whatever it's hooked up as.

_________________
Когда хочется кушать – съешь всё.
Голод не тётка!

I had my suspicions about mounting, assuming it could attempt to write if it mounted, thanks for confirming.

Now I just have to wait the two weeks for the drive to catch any files I can grab, and the psu molex to sata power adapter.

Thanks for your help, I'll definitely post results.

I believe the program I started was PhotoRec. I remember it was looking for file flags, but you mentioned about defragging. I did not defrag, but this drive was only receiving new files, never deleted anything off it, never a download/torrent target. It only had complete files added, if that makes a difference.

Depends what spaces were there and if the OS had need to shuffle things.

With apologies to yourself and other technicals this is the best i can manage for my completely non-technical customers should they really require me to cure their insomnia as well as explain filesystems in a non technical way.

Fragmentation for the lay person.

So, the library throws out a big fat encyclopedia they no longer need, (deletes a file) or fancies moving it just because it can. (auto defrag, and maybe stops half way thro)

Therefore there is a big fat space left. (slack space)

Someone then brings a new book in, and half fills the space. (installs a program or saves a doc/photo/music file)

The librarian shakes her head as there is only a thin space left.
Heres the rub:
The librarian will find an appropriate space, shuffle the books up thus expanding the space for the new book, or put the book at the end of the shelf where there is sufficent space.

The book remains intact and can be found by anyone else seeking the complete book. The pages remain between the covers (header and footer)

Your OS, however, finds the (best fit or worst fit ) next available space and may decide to rip as many pages out of the book as is necessary for it to fit, or just tear the entire book up, put shreds of paper anywhere and index where they are.

Anyone requiring the book (file) now, needs to check the index (file table) in order to extract the book as a collection of disparate pages which are then temporarily reconstructed in memory.

Without that index you may find the cover, (file carving header and footer) because thats where the book title is held, and the first few pages before the great rift, but thats about it. The shreds could be and are usually everywhere else outside the covers.

The smaller the books are (small filesize), the more chance you have of recovering the complete book without a library index as the gaps that are left can usually accomadate pamphlet and leaflet type books. (small .txt .jpg .doc .xls)
The larger Encyclopaedia Britannica (video files) are so large it is almost impossible to find a shelf long enough to store it unbroken.

In terms of data recovery, the gold mine is the recovery of the index.
With that you don't have to image the entire library in order to find the book you require.
This is also why hardware solutions like the Deepspar and Atola imagers are superior to just dumb dd type operations. You can extract by filename or other parameters often saving days of copying zeros interspersed with bad sectors in order to get a "full image" (forensic cases excluded)

Phew.
Apologies to pro's here for bending things a bit and inconsistencies.
Just thought it may help some ppl out with the concept.

Kern

_________________
Когда хочется кушать – съешь всё.
Голод не тётка!

ddrescue is better. It understands how to work around bad media.

_________________
A backup a day keeps DR away.

Thanks for the replies.

Let me know if I've missed something in the manual, or if I need to start a new thread/go to MHDD forum -

Trying to run MHDD

Power on, WD20EARX detected by BIOS, tried both Auto and Large Block in BIOS

Boot to UBCD on USB

Select HDD/Diagnosis/MHDD32 (the 4.6 one)

Select Without SCSI Support (doesn't make a difference to below woes)

Three choices - 1. 1F0h (primary controller) 3. 170h (secondary controller) 5. 100h (PC-3000 board)- No mention of my WD20EARX, which I believe is related to Autodetect and mhdd.cfg. I wonder if this is a trick question to weed out noobs, I am tempted to enable to see what happens.

1 and 3 lead to screens with all the status register lights out, and F2 gives a "device not ready" error, 5 has red ERR, blue INDX CORR DREQ DRSC WRFI DRDY and green BUSY, F2 thinks for a moment then gives Device Not Ready.

So I am not getting past the front door.

Questions -

Where are the logs/screenshots going to go? I'm running off USB, ideally, they would be written to the USB, or I might be able to set a log location. I also expected to see some way of telling it to copy the log file I even made a second partition on my USB to dump things in case I found such a setting. I would like to be able to post the screens rather than trying to explain.

Why isn't the drive getting identified anywhere? I understand no autodetect (I think) but I'm second guessing my understanding when I read about setting jumpers to master and IDE and all. MHDD makes no mention of SATA.

Thanks for any advice.

Configure your SATA controller in BIOS for legacy IDE mode.

_________________
A backup a day keeps DR away.