Best app for partition recovery

[Jim]

Any suggestions for a decent app to recover lost partitions?

[Flasherly]

On Sun, 16 Feb 2014 12:51:44 +0000, Jim
<luckyjim...@yahoo.co.uk> wrote:

>Any suggestions for a decent app to recover lost partitions?

Easeus

[Mr. Man-wai Chang]

On 16/2/14 8:51 PM, Jim wrote:
> Any suggestions for a decent app to recover lost partitions?

SUGGESTION:

Don't try to do two things at the same time and with one single tool.

First, restore the partition table.

Then you restore the data inside the partition.

Whatever you do, **DO NOT** directly modify the content of the hard disk
in question. Files restored should be copied to another hard disk. Tools
like Recuva does this.

--
@~@ Remain silent. Nothing from soldiers and magicians is real!
/ v \ Simplicity is Beauty! May the Force and farces be with you!
/( _ )\ (Fedora 19 i686) Linux 3.12.9-201.fc19.i686
^ ^ 23:18:03 up 2 days 3:05 1 user load average: 0.20 0.13 0.08
不借貸! 不詐騙! 不援交! 不打交! 不打劫! 不自殺! 請考慮綜援 (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa

[Mr. Man-wai Chang]

On 16/2/14 11:26 PM, Mr. Man-wai Chang wrote:
> Whatever you do, **DO NOT** directly modify the content of the hard disk
> in question. Files restored should be copied to another hard disk. Tools
> like Recuva does this.
>

A simpler way is to clone the hard disk first, then use recovery tools
on this copy of the hard disk.

--
@~@ Remain silent. Nothing from soldiers and magicians is real!
/ v \ Simplicity is Beauty! May the Force and farces be with you!
/( _ )\ (Fedora 19 i686) Linux 3.12.9-201.fc19.i686

^ ^ 23:24:02 up 2 days 3:11 0 users load average: 0.05 0.06 0.06

[Paul]

Jim wrote:
> Any suggestions for a decent app to recover lost partitions?

There are "two levels of lost".

If the MBR is damaged or overwritten, then the individual
partitions can no longer be found. For that, you use TestDisk.

TestDisk computes new values for the partition table, based
on the file system headers it finds. The method is problematic,
in that a deleted partition can be relocated. And, the tool
doesn't necessarily check for or detect overlap in the
definition of partitions. It's up to the human operator,
to judge whether the new, proposed, MBR value is any good.
So in a sense, it's far from fool proof. You have to eyeball
the results, and decide whether "the disk originally had
two partitions, a small one and a big one".

http://www.cgsecurity.org/wiki/TestDisk_Step_By_Step

At the next level, is Disk Management shows the partitions
OK, but one of the partitions doesn't have any type information.
Perhaps the partition will no longer mount. In that case,
you need something to recover the partition.

The only free one I know of, with initial NTFS support,
is Drive Rescue. Which is ancient.

http://web.archive.org/web/20070101070056/http://www.woundedmoon.org/win32/driverescue19d.html

This is a screen shot of Drive Rescue at work.

http://imageshack.us/a/img822/1299/evie.gif

*******

For forensic work, you need two hard drives. One
hard drive, holds an exact sector-by-sector copy
of the damaged disk. The second spare hard drive,
is where DriveRescue is going to save the recovered
files.

*******

There are many $39.95 recovery softwares. And
some of them are designed to show the file names,
as a teaser, to prove the tool can find them.
(That's the "trial mode" of the product.)
You then pay your $39.95, and see whether the tool
actually works and can recover them. I have no
experience with any of these, because I don't
have N * $39.95 to waste :-)

Good luck,
Paul

[Jim]

Hi guys thanks for your replies so far i'm in a real pickle here.
The drive is a new WD3TB Black, however it is only showing up in recover
software as 747GB, i need it to be able to show me what is on the whole
3TB drive any idea folks?
I love these fat drives but when you loose em it's a nightmare.

Jim

[Paul]

Jim wrote:

> Hi guys thanks for your replies so far i'm in a real pickle here.
> The drive is a new WD3TB Black, however it is only showing up in recover
> software as 747GB, i need it to be able to show me what is on the whole
> 3TB drive any idea folks?
> I love these fat drives but when you loose em it's a nightmare.
>
> Jim

Piece of cake. I've seen this, but it's a function of
the OS you're using. I get similar symptoms in Win2K.
(I have a 3TB WD Black.)

It also depends, on what you used to prepare the drive.
Western Digital provides a free copy of Acronis TIH, suitable
for installing the Extended Capacity Manager. Only problem is,
at some point in the past, I installed some other Acronis
software, it installed a driver, and the driver was not
removable. So the new software, could not install its driver.
This complicates the process of getting the
Extended Capacity Manager working.

I can promise you, that the data is still accessible. I was
able to access the upper 747GB, using a loopback mount in
Linux, using a little-known offset parameter. That allows
me to get to the partitions that might be hiding up near
the end. But to do that, I had a fairly good idea, what
the offset should be. It only took a little bit of searching,
by taking snapshots with "dd", to figure out where the file
system header was.

I doubt one of the older data recovery programs is going
to be entirely happy working on your problem. So perhaps
you can provide a few details of what you've tried, and
I'll see if I can jog my memory as to what I did to fix it.

I've worked with the Extended Capacity Manager twice,
and both experiences were horrible (blinding rage horrible).

The Acronis TrueImage Cleanup Utility is here. It
can remove the old driver, if one is present. The Acronis
staff thought you could not successfully write one of these,
but a user in one of their forums, showed them how.
(If the driver is uninstalled in the wrong order, it
would cause the OS to blue screen.)

http://kb.acronis.com/content/34876

A standalone driver package exists. When I installed
this in Windows 8.1 Preview for example, I was able
to use the entire 3TB of disk, just as I see it in
WinXP today. But the required metadata was already on
the disk, for that driver to use.

https://kb.acronis.com/system/files/content/2013/01/38937/virtualdisksetup2013.zip

But what that doesn't do, is it doesn't do the
Extended Capacity Manager step. ECM writes a 256KB
chunk of metadata, at the 2TiB mark. The upper virtual
disk, appears just after that point. And both times I
worked with ECM, it *refused* to allow me to click the
button, and install. And I didn't keep careful notes,
with all the flailing I was doing, of what fixed it.

So bottom line is:

1) Your data isn't lost. It can be accessed from Linux, for
free, with a bit of work. The upper 747GB or whatever,
is treated like a big bitmap. The -o loop option in the
Linux mounter, allows a file system to be mounted on a
mount point. Even though Linux would not normally allow
access above 2TiB on an MBR disk, the offset in the loopback
mounter supports 64 bit numbers.

2) Depending on OS, the OSes behave stupidly, even when you
do things mostly right. I've tried cranking down the
first 2TB partition by small amounts, to get them to behave
better. But I think what was really pissing off the OS,
was cylinder offsets (WinXP style) versus megabyte offsets
(Windows 7 style). The Extended Capacity Manager, was using
one style on the lower disk, and another on the virtual disk.

Acronis should be soundly whipped, for the mess they made there.
Not a pleasant experience at all, and double the work for
me that it should have been. Thank God they made that
Cleaner utility, or I'd still be swearing and tearing
out my hair!

Paul

[Jim]

Hi Paul thanks for the reply, boy is this a can of worms i wish i had
never opened and stuck with my smaller 1TB Blacks.

OK when i started using these 3TB drives i was coming up withthe 2.2TB
issue and suggested to use GPT or something like that but what i didn't
like was at the begining of the drive was a smaller 128mb hidden
partition so i wanted to find a way round it and format the drive
keeping it nice and clean and i used so many differnt apps to get it the
way I wanted i thinkn i used a free app from Paragon called "Paragon
partition manager free edition" but i can't be 100% to be hobnest on how
it was formated it could have been a number of way. i was going to
reformat it using GPT but thought better of it, as yet i have wrote
nothing to the dud drive. I know my data is there but getting it is not
straight forward.

I am running a new instal of win 7 U x64 to be honest i should be able
to get my hands on any of the latest software most offer shareware with
option to upgrade to recover so i'm happy to do that, i'm willing to buy
anything to get my data back.

Jim

[Flasherly]

On Sun, 16 Feb 2014 22:06:52 +0000, Jim
<luckyjim...@yahoo.co.uk> wrote:

>I am running a new instal of win 7 U x64 to be honest i should be able
>to get my hands on any of the latest software most offer shareware with
>option to upgrade to recover so i'm happy to do that, i'm willing to buy
>anything to get my data back.

Knowledge can be in a currency at times difficult to convert from
experience: what irony serves, in this case, that a 3T drive wouldn't
offhand avail itself any less readily to a larger function and
capacity of storage, at $30US a T, comparatively to $70 for an initial
entry offer to platters (or, hypothetically, in silicon SSDs for
eminent price-drops at usually some inevitable future point).

Right now, I've two docking stations limited to less than 2T drives,
and only one DS able to handle 2T. By formatting/partitioning my
physical 2T drives, at less than 2T, from a either a MB BIOS, or on
that 2T compliant DS, say, for being within a capacity for the
less-than-2T docking stations to indeed recognize (for established
storage data, copying/writing purposes).

Theoretically. Not sure I've back tested that line for a stratagem;-
most certainly not at 3T, as I haven't that large of a drive yet.

At USB2 10-12MByte transfer speeds, a travail of absurdity, perhaps
over days to fill sizeable 3T drives.

Reminds me of an early 20 Megabyte drive on dedicated controllers
before hard drives were established for a BIOS function. To test
20MBytes of archived data in compression formats, before ZIPS, on a
4.7MHz 8088 or 6MHz NEC V20 necessitates 48 hours.

Indeed, how time literally flies when you're having fun, yes?

[Paul]

According to this, the partition type on the protective MBR would be
0xEE if you had set it up GPT.

http://en.wikipedia.org/wiki/GUID_Partition_Table

You can use PTEDIT32, to look at the MBR partition table. Even if
it's the "fake" "protective" MBR, you can still look at it to verify
you did set it up as GPT. You unzip that first. Then, right click
on the ptedit32.exe program, and select "Run as Administrator". The
program returns an Error 5 if you attempt to use it without the
Administrator authority (since it accesses the raw disk).

ftp://ftp.symantec.com/public/english_us_canada/tools/pq/utilities/PTEDIT32.zip

Once you've got some positive feedback, as to what it is,
you can try TestDisk, as a confidence builder. The idea
would be, to see if TestDisk sees GPT or not.

http://en.wikipedia.org/wiki/Testdisk

"TestDisk recognizes the following disk partitioning:

...
GUID Partition Table <=== GPT
"

Note that, most of the time, you get an "opinion" from TestDisk,
and don't accept it's offer to write anything back. If you
need to quit TestDisk, you can press control-C to stop it.
It would probably take a long time, for it to scan the whole
disk on 1MB boundaries or something, looking for file system
headers. The reason for not immediately accepting an offer
to write a new disk header, is because it doesn't do enough
checks and balances. But still, it can be educational, to
see what it is capable of digging up.

TestDisk can display the file names in a partition it finds.
Which would be another confidence builder.

http://www.cgsecurity.org/mw/images/List_files.gif

( http://www.cgsecurity.org/wiki/TestDisk_Step_By_Step )

*******

The article on "GUID_Partition_Table" says:

"GPT also provides redundancy, writing the GPT header and
partition table both at the beginning and at the end of
the disk.
"

That suggests to me, if something has "gone missing", it's
the file system header in the partition itself that
got damaged somehow.

You might also check Event Viewer, to see if the
software that attempts to mount file systems, has left
any error message(s) about what it found.

Using CHKDSK, would be reserved for the home stretch,
when the partition is mountable. CHKDSK is a double-edged sword,
in that, if the disk is health, it's probably worth running.
If the disk has problems with read/write or mechanical problems,
CHKDSK can make things much worse than when you started. CHKDSK
has been known to entirely trash a partition, when the disk itself
is sick or an IDE cable is slightly loose.

Paul

[Jim]

Afternoon Paul, well this is looking more worrying as it only ever seems
to show up as 750 odd GB, having ran PTEDIT32 and testdisk i took some
screen shots and stopped because testdisk gave a message saying not to
go any further if disk size was reported wrong which it is, below are
the links for screen shots.

http://imageshack.com/a/img716/2622/34nm.jpg
http://imageshack.com/a/img593/9896/gk7c.jpg

I was being quite optimistic last night and thought if this went well i
will not have enough spare space to take stuff off so i bought another
3TB Black last night, looks like i might have been kidding myself :(

Jim

[Paul]

Yeah, my problem is, I don't understand what the OS is doing
in this case.

Your first picture confirms GPT. That's the protective or fake
MBR we're looking at. And the size field corresponds to 2TiB
for the fake (as large a 32 bit unsigned number as possible).

The second picture, looks like what I might see on Win2K.
And I don't understand what Win2K does in that case. I mean,
I could understand, if some math was done with 32 bit unsigned
arithmetic, that is where the 747GB number is coming from.

I guess I could ask silly questions, like is the current OS
viewing the drive, the same OS as prepped it in the first place.
I checked the Wikipedia article, and GPT is supported in
Vista/Win7/Win8. As well as Ubuntu >8.04 (it's up to 13 now).

The BIOS may play some part in this. It probably reports the
size of the drive, to the OS at some point. Choices for
information presentation, would be a CHS tuple (fake geometry)
and LBA size (let's hope that is a 64 bit number). At that point,
I kinda lose my theoretical footing, like what might convince
the BIOS to screw up.

So a possible solution there, would be to write down any
custom BIOS settings, then try a Clear_CMOS. That procedure,
using the jumper, is only to be done with the computer
unplugged. That's a conservative practice, that prevents
bad motherboard design, from causing damage. Some motherboards,
the CMOS jumper shorts something that can burn a particular
component. The user manual would probably say, to remove
power from the computer. In years past, I've found a number
of manuals with mistakes in that section. Including a manual
that encourages you to leave the jumper in the wrong position.
The purpose of clearing CMOS, is to have the BIOS regenerate
the datafill of the CMOS RAM.

Some BIOS have a "Load Default Settings" option. While you could
try that, it may not clear every byte in the CMOS RAM. But may be
sufficient in this case (no reason to assume the un-defined areas
of CMOS, cause the symptoms). After Load Default Settings, doing
a Save, then you go back (without booting the OS)< and enter
the custom settings you wrote down earlier. For example, on my
older P4 machine, if I clear CMOS, I have to remember to disable
the Promise Controller, and correct a couple other settings
that don't have good default values. On the machine with the
sound card, I might disable motherboard audio. Just trivial
stuff. Several of my machines might have custom RAM timing,
and I'd have to put that back.

I haven't been able to figure out, where else the geometry
information is coming from.

So some quick tests would be:

1) Boot an alternate OS, and see if the same size value is
visible. If the entire GPT is visible and working, now
you're ready to transfer to another drive.

2) If the alternate OS suffers from the same problem, try
to Clear CMOS or Load Default Setup. Reloading settings
as before. Most IDE disk setting, would be set to LBA,
rather than Large or CHS. Each alternate setting, can have
an effect on the geometry information the BIOS passes to the
OS.

Other than that, because I don't understand what party
is playing with the geometry, I don't know what knob to turn
next. I mean, if I knew, I would have fixed this on my
Win2K installation, so that my 2TiB drive would stop showing
up as the wrong size. Fortunately, nothing bad has happened
with all those size issues. Only if a partition "spans" a
capacity barrier, do you run the risk of corruption.

*******

This page gives some idea, of the sources of geometry info.
(Click the "X" to dismiss the popup they overlay on the good stuff.)

http://support.novell.com/techcenter/sdb/en/2004/07/fhassel_edd.html

BIOS: CHS = 10/8/8 bits 1023/255/63
LBA = 64 bit number

Stored on disk itself (MBR?) = CHS, different than BIOS.
Using fake geometry here, allows
changing disk alignment (like on SSDs)
I've done this once, just for kicks.
Changed sectors to 56 or something.

You can see in this example, that Linux appears to have an
interface to the CHS stored in the MBR. You can pass values
on the command line.

http://manpages.ubuntu.com/manpages/precise/en/man8/fdisk.8.html

OK, it just occurred to me, the PTEDIT32 has a display of
on-disk CHS value. Yours is 97451/255/63. Doing the arithmetic
(roughly, by ignoring whether I'm supposed to be adding one to
one of those fields or whatever)

97451*255*63*512 = 801,561,761,280 <--- 747GB

Oh, dear. So that's where the size is coming from. OK, I'm going
to have to shut down the PC, and plug in my 3TB drive now, for
comparison. Of course, it isn't GPT, so the comparison isn't exactly
valid.

Back in a moment...

Paul

[Paul]

Paul wrote:

> Jim wrote:

>> Afternoon Paul, well this is looking more worrying as it only ever
>> seems to show up as 750 odd GB, having ran PTEDIT32 and testdisk i
>> took some screen shots and stopped because testdisk gave a message
>> saying not to go any further if disk size was reported wrong which it
>> is, below are the links for screen shots.
>>
>> http://imageshack.com/a/img716/2622/34nm.jpg
>> http://imageshack.com/a/img593/9896/gk7c.jpg
>>
>> I was being quite optimistic last night and thought if this went well
>> i will not have enough spare space to take stuff off so i bought
>> another 3TB Black last night, looks like i might have been kidding
>> myself :(
>>
>> Jim
>

> OK, it just occurred to me, the PTEDIT32 has a display of
> on-disk CHS value. Yours is 97451/255/63. Doing the arithmetic
> (roughly, by ignoring whether I'm supposed to be adding one to
> one of those fields or whatever)
>
> 97451*255*63*512 = 801,561,761,280 <--- 747GB
>
> Oh, dear. So that's where the size is coming from. OK, I'm going
> to have to shut down the PC, and plug in my 3TB drive now, for
> comparison. Of course, it isn't GPT, so the comparison isn't exactly
> valid.
>
> Back in a moment...
>
> Paul

OK, my 3TB disk uses the physical+virtual Acronis method.
The lower portion shows up as a 2TB disk, the upper, as 747GB.

The 2TB shows this for CHS

243201/255/63 = 243201*255*63*512 = 2,000,396,321,280

So that's consistent with the partitions defined in the lower MBR.

The upper part of the drive, has its own MBR, which the filter
driver makes to appear as a separate drive. And that one, not
so coincidentally, is

97451/255/63

as that's the left-over space being defined as a virtual drive.

*******

Now, in these, I don't see where any disk geometry for the whole
disk is stored.

http://thestarman.pcministry.com/asm/mbr/PartTables.htm

http://thestarman.pcministry.com/asm/mbr/GPT.htm

http://thestarman.pcministry.com/asm/mbr/W7MBR.htm

The last article, makes it sound like the partition
entries are fashioned, based on BIOS reported geometry.
But of course those numbers can't change, if the BIOS has
a brain fart and starts doing something different. However, the
total size of the disk can be passed as a different value, leading
to the truncation you're seeing.

So my best guess is, the BIOS is doing this somehow.

The EDD spec, doesn't look like it has any geometry in it,
fake or otherwise. And yet the Linux article I was looking
at, was claiming it was getting it from there.

http://t13.org/Documents/UploadedDocuments/project/d1386r0-EDD.pdf

The hwinfo output in Linux here, shows an EDD geometry. And maybe that's
related to what we're seeing in your case.

http://unix.stackexchange.com/questions/4561/how-do-i-find-out-what-hard-disks-are-in-the-system

I'll boot over to Linux now, and have a look at hwinfo.

*******

Well, stumped as usual. Still don't understand this stuff.

All I can suggest at this point, is to think carefully about
any changes made recently. Like BIOS changes just before
it could no longer mount.

Paul

[Flasherly]

On Mon, 17 Feb 2014 11:46:28 -0500, Paul <nos...@needed.com> wrote:

>All I can suggest at this point, is to think carefully about
>any changes made recently. Like BIOS changes just before
>it could no longer mount.

Why I like "the idea" of coming off a docking station, so far as
setting up a drive. My largest DS firmware support being 2T, I've
certainly aways to go [w/out a 3T drive to my name], although being
USB, there should be a layer [of USB transpostions, i.e., from within
the station's firmware] non-specifically addressed to the MB's BIOS;-
interesting distinction how that differs from an added SATA port my
stations haven't.

Anyways, slap up that baby upside any ol' USB2/3 port, a baseline
compliancy, and the given software/OS originally used to establish the
drive's geometry [reported to the OS] theoretically should be good to
go to work.

[Paul]

That's a good point.

The right USB disc enclosure, might just do it. Select a
USB enclosure where it states in the advert, that it handles
a 3TB or 4TB disk. I think some of the USB3 ones advertise such,
and the USB3 one can be plugged into a USB2 port. I plug my USB3
flash key into USB2 ports all the time.

Paul

[Flasherly]

On Mon, 17 Feb 2014 14:08:28 -0500, Paul <nos...@needed.com> wrote:

>
>That's a good point.

Amen. Two were given to me (for HD purchases - 1.5T limits), the
other ran $20 (2T limit). All Rosewill.

Interesting, too - is my present BIOS(es) - some I'm not exactly sure
which, over three MBs, haven't that capacity to handle 2T drives,
either. Simple matter, slap it upside in a docking station, run
Easeus to partition it out (via the station's firmware), slap that
baby into the cage array, off the incompatible MB's SATA controller,
and it apparently can't tell the difference. Comes up looks/functions
regularly.

Been awhile since messing with them, and I'd have to double check back
for mistakes, but believe I've perhaps pulled off such. (There's also
a semi-wacko PCI HD controller, SYBA or such, in one that also took a
docking station pre-formatted drive;- drops speed occasionally and is
dog slow with optimizing with Udefrag3 - UltimateDefrag3.)

Horseshoes and hand grenades, modern day marvels of wonder, for living
on "borrowed time," between all this old equipment and the kludge I
put them thru.

[Jim]

Well i do have an IcyBox IB-RD4320STU3 so i will give that a bash later
and see if i can get any joy from that route.

Jim

[Jim]

OK well after putting the drive in the enclosure and booting it up in
win7 it was not showing up in windows explorer at all so i went to disk
management and it wasasking me to initialize the disk with eityher MBR
or GPT at this point i bottled it as dont want to write anything to the
drive yet, i do have my new 3TB black here right now so i'm going to try
to clone the broken 3tb onto the new one so i can at least have a back
up of it if i do something silly, I may look at ebay and get a different
kind of docking station as to be honest the icybox i would not class as
a dockling station myself.

Jim

[Jim]

Well thought the best thing was to put new drive inline and let windows
format it the way it wants to but oh no, went GPT route and now it is
now showing as 764.39GB, guys i'm begining to wonder if my motherboard
could be the issue here, i dont know why, sure it's an older model Asus
P5E3 Premium @ wifi
https://www.asus.com/Motherboards/P5E3_PremiumWiFiAP_n

Anyone got a gun?

Jim

[Jim]

I have been thinking about this and was wondering could it be the drives
themselves? Reason i'm think this is i have bought the newer version of
the drive the AF format (WD3003FZEX instead of older WD3001FAEX drive) I
was told there was no real difference but who knows?
http://wd.com/en/products/products.aspx?id=760

Jim

[Paul]

Nope. I suggested Clearing CMOS as a cure.
Or, using Load Default Settings in the BIOS,
as an easier substitute. You will then need
to load any custom settings later (disable
peripheral chips, load custom RAM settings,
overclocks etc). If you've never ever used
custom settings, accepted the default IDE emulation
mode etc., then you'd have virtually nothing
additional to do.

There's no geometry in the MBR sector. There is
definitely geometry in the four primary partition
entry area, but that doesn't influence the overall
drive size. The BIOS can read the max_address out
of the drive, and convert it to a fake CHS. And in
Linux, it would appear the parameters to fdisk, are
"fakes" to influence how the primary partitions
are set up, rather than overriding anything in
a permanent way. I think I may have used that,
so I could test WinXP on a 56 sector alignment.
(Prep disk in Linux, and transfer files over later.)

So all I can figure, is the BIOS is doing it. If
the 3TB drive had never worked properly, I'd just
say "toss the motherboard". But it did work at
one time, so keep the motherboard. Try clearing
the CMOS (with all power removed - unplug the computer).
The green LED on an Asus motherboard, should not be
lit while using the CMOS jumper. Wait at least 60
seconds after unplugging, before making any changes
inside the machine. That's to give time for +5VSB to
"drain". The computer continues to draw current from
it, which is why it will drain. Someone suggested
pushing the power button on the front, to encourage
draining, but I consider that to be overkill. The
supply might not act sanely, while such a transient
happens.

Paul

[Jim]

OK that's fine i'll do that this afternoon Paul, I have not reset the
bios in yonks and my bios is fairly standard apart from ACHI and
disableing sound etc so no overclcoking done here, i prefer stability
over performance, I'll post back later with an update.

Jim

P.S. Thanks for staying with me.

[Flasherly]

On Tue, 18 Feb 2014 12:22:05 +0000, Jim
<luckyjim...@yahoo.co.uk> wrote:

> I may look at ebay and get a different
>kind of docking station as to be honest the icybox i would not class as
>a dockling station myself.

http://www.newegg.com/External-Enclosures/SubCategory/ID-92

Haven't looked much at them, earlier, last nite, for couple quick
reviews out of curiosity - say, the top-positioned Orico, which has 3T
firmware support for USB3/eSATA, so should be decent speeds (some
reviews were getting it, others not).

There's also going to be other 3T makes, of course, more money - not
that Orico's $38 is really cheap (potentially questionable make to
trust relatively expensive HDs, though $20 is all I've paid for my
Rosewills).

Sort them by popularity/most reviews and dig into the user
experiences, I suppose, if you're interested in the external route.
Beats the "leap of faith" on Ebay if not especially up on what make is
offered, (such as offbrand Singapore equipment imports), specs or a
broader reception for satisfied end users.

Appears likely that 3T firmware is going to be in the minority.
Conspicuously so when Thermaltake, a premier docking brand, I don't
offhand readily see positioned against lesser competitor brands among
3T offerings.

Without a solid MB USB3 performer mated correctly, there's only so
much worth in a futility involved in these things -- USB3 PCI cards
for docking stations and such. Get's expensive at some point, more
worth in just buying a good MB, in the first place;- all I've got in 3
docking stations, a PCI SATA 2-port controller is about $30, total.
Gets me by with a stack of 1.5T and 2T drives on 3 older, single-core
and earliest dual-core MBs. Slowly, needless to mention (I've a 115V
fan for sustainted "docking work" -eh.)

[Jim]

Well bios is still showing it as 801.6GB despite the fact i can see it
as 3TB in various boot discs i have picked up along the way (hirens
etc), I have scanned WD forum and tried what is mentioned here
http://community.wd.com/t5/Desktop-Mobile-Drives/2-WD30EZRX-and-WIndows-7-64-bit/td-p/370397

but still coming up with 750 odd GB options even tried linux partition
app (sorry can't remember name it's 3am here) but coming into windows
still gives me 746.52GB Unallocated as i have not even bothered
formatting it.

I have a few more htings to try someone has mentioned about my bios may
be truncating (spelling) my drive so i have to run some tests on that
later but it's gone 3am here (London) and i'm dead so going to call it a
night I will post with some results later for you to have a nose over.

Nite all

Jim

[Jim]

Well havng slept on it i thought i'd try something different and that
meant i'd hook up the new (empty) 3TB to my Adaptec RAID 1430SA
controller which i know supports 3TB drives, so booted into windows went
into Disk management and formatted perfectly, so then shut down and put
the new 3TB back onto the motherboaurds SATA controllers to see what the
bios would say and it's coming up again at 801.6GB so i'm sure some of
my problem come from my motherboard.
So i'm now going to hook up the original drive and see if i can image
that but to be honest i'm not sure if i can but worth a go, then i'll
hook that drive up to the raid controller and see if i can do any
recovery work on that.

Jim

[Paul]

As I said before, if all else were to fail, you can still locate the
start of the NTFS file system, and loopback mount it in Linux. All
the data should be there. Then, in Linux, you'd format the new
drive, and transfer the data over.

The loopback mount (-o loop), with appropriate offset, is a way to
pick off a file system when the header no longer points to it.

GPT stuff... mumble mumble NTFS_header ... NTFS_data
^
|
Set offset to here.

But both drives should still foul up in Windows again. Until the
mystery of where the errant geometry is coming from, is solved.

When I used that technique in Linux, I was able to mount the NTFS
partition above the 2TiB mark. Proving that offsets above 2TiB, work.
Data transfer was pretty slow (10MB/sec, versus the normal 135MB/sec
or better), but it's better than nothing.

If you fired up Linux and the drive was correctly identified
from the start, then you'd suspect that somehow, a drive overlay
got loaded in Windows. This is another thing that can work
under the hood. I don't know if they make a DDO to solve large
drive problems or not. You'll notice "Kroll Ontrack" is
mentioned in this article, and Ontrack software is typically provided
on disk manufacturer web sites (while they're still licensing and
paying Kroll for it).

http://en.wikipedia.org/wiki/Dynamic_Drive_Overlay

Have a look at your Add/Remove, and see if you can figure out
what software was added to your computer, around the time
the geometry went to hell.

Paul

[Jim]

OK Thanks Paul trouble is what i know about Linux could be wrote on the
back of a postage stamp, and room to spare come to think of it, any
pointers on what iso is best to grab and even a link for it, and then i
can look at going down that route as well, from my point iof view i have
nothing to loose.

Jim

[Paul]

The one I keep on a USB key right now is Linux Mint (Mate Interface).
I keep what's called a persistent home directory (casper-rw), and that
saves my results from session to session. The rest of the files for
booting the OS, are read-only.

The My Computer on the upper left, has the partitions showing.
The lower-left corner has a menu with reboot stuff and a Terminal session.
Command line things are done from Terminal.

http://upload.wikimedia.org/wikipedia/commons/a/ac/Mint_15_RC.jpg

http://www.linuxmint.com/download.php

MATE 32-bit 64-bit An edition featuring the MATE desktop Multimedia=Yes

The 32 bit one should be good enough. Sometimes, things like 64 bit browsers
aren't available or the 64 bit plugins don't work. The 32 bit is safer in terms
of getting something out of it.

The ISO9660 file will be around 800+ MB. They make them a bit bigger
than CD sized now. Since I put mine on a USB stick, it doesn't cost me
anything in terms of non-renewables.

It's a pig at boot time. It's probably a 50% longer boot time than Ubuntu,
and I don't know exactly why. Even when a LiveCD is bloated, they don't
have to load every byte on the CD/DVD. But it sure feels like it in
this case. Even on the USB stick, at 30MB/sec, it's slow. Like, even
the menu in the lower-left corner, is slow to "wake up" and become
ready. The record for a free OS booting is 5 seconds, but this one
is a big loser in that contest. More like 3 minutes.

In Terminal, you preface commands with "sudo". It's like a "Run as Administrator".
For example, to get disk geometry, you could try in terminal

sudo hwinfo --disk

and it would scroll down the screen. The Terminal program preferences
can be set to infinite scrollback, so that no output from commands is lost.
By default, a Terminal session doesn't usually have that set large enough.

At shutdown, the OS will dismount partitions. I usually dismount
all the optional ones manually, but that's just a force of habit.

In Terminal, a command like (diskfree)

df

shows the mounted file systems.

And something like

sudo umount /mount/mint/<partition_identify>

would dismount a partition.

The manual pages are found by typing

man umount

and typing q to quit the manual page.

Anyway, the LiveCD is a useful thing to keep around, if you
run out of options otherwise. I like to compare hardware
behavior between OSes, to determine which are OS problems
and which are hardware problems. Even if triage there, isn't
of much use (i.e. no CHKDSK available). There is a chkdsk
type option, but all it does is set the dirty bit, forcing
Windows, the next time it is running, to fix up the partition.
There is no code in free Linux, to repair NTFS partitions.
A commercial Linux product is available, for around $100 or so
(dream on). There are certainly free software developers who
know NTFS inside-out, but nobody wants to write a CHKDSK.

Paul