More USB Disk Drive Woes

Last week I wrote about my problems with a Lacie Quadra d2 USB drive.

Short version: the drive became unrecognized. Most tools wouldn't recognize the device at all. Disk Management gave me the opportunity to Initialize the device but that always failed with an Incorrect Function error.

More details here:

http://www.dpreview.com/forums/thread/3532693

One week later a second USB drive has failed. This time a WD MyBook but the failure symptoms are worryingly similar.

Once again, Initialization of the device fails; this time it says the device is write protected (although DISKPART says otherwise). The WD Diagnostics fail almost immediately with "Too many bad sectors".

Interestingly, Windows and the WD tools report the device as an "Elements" drive when it is in fact a "MyBook".

In both cases, the symptoms are consistent with a corruption of the basic device parameters (model no, geometry and the like).

Both drives exhibit the the same problems on my desktop and laptop machines. Using different USB cables and ports doesn't help. Also, both drives are normally powered via a good quality UPS with surge and spike protection etc. so I don't think these issues are power related.

I don't like coincidences like this and worry about some common cause and the possibility that my other USB drives are at risk :-{

I have checked my systems with Malwarebytes, Emsisoft, and Spybot and no malware has been found.

The WD drive does contain useful data but I do have one other copy (this is bulky but low value data so I don't have more copies). It is also under warranty so I certainly have the option of RMAing the thing but doing so will rule out any further analysis or testing.

My biggest worry is the vulnerability of my other half dozen USB drives.

Any thoughts or suggestions welcome!

Have you updated your BIOS firmware and made sure the USB chipset drivers are up to date in case it's a hardware/driver issue of some type causing it)?

Have you tried the same drives on another PC?

How about from a Live Linux distro of some flavor running from DVD or CD to see if symptoms change (which would help to rule out some type of root kit or boot sector malware causing it that may be missed by scanners when booting into an already infected OS).

Just to rule out malware, I'd probably try a few Linux based AV tools (so you're booting into a clean OS to run the scans).

One easy way to do that is using a USB Flash Drive (or a memory card in a USB attached reader) that you can boot into. Just use something like YUMI to install a variety of different AV products on the same flash memory device.

Scroll down in this post for details on that approach:

http://www.dpreview.com/forums/post/42554161

A lot of the major AV products have linux based rescue disks. So, I'd probably scan your system with some of them, just to be safer (as malware can more easily hide from AV software if you're booting into an already infected OS).

Of course, it may just be a coincidence (as drives do fail from time to time). This is Windows 7, not Windows 8, right?

If Windows 8, then you may need to take some extra precautions like disabling fast boot (especially if setup in a dual boot config with Linux).

Otherwise, Windows 8 can think the drives are in the same state they were in at the time you shut it down last. See this article for more info:

http://www.h-online.com/open/featur...8-Fast-Startup-puts-data-at-risk-1780640.html

--
JimC
------

corrupted, and can not be accessed? so regarding the first one, were you able to reformat and create a new partition and then check the health of the drive? did it work correctly?

Member said:

malch wrote:

Last week I wrote about my problems with a Lacie Quadra d2 USB drive.

Short version: the drive became unrecognized. Most tools wouldn't recognize the device at all. Disk Management gave me the opportunity to Initialize the device but that always failed with an Incorrect Function error.

More details here:

http://www.dpreview.com/forums/thread/3532693

Click to expand...

Trying to format the drives with another computer to format them.

You said in previous post the drive was pulled out. Could be a USB problem. Try another computer and cable.

Do you have a magnetic field problem near your computer or drives, has the location of either changed recently.

Is your power surge protected, brown out protected? (good power)

digitalshooter said:

digitalshooter wrote:

corrupted, and can not be accessed? so regarding the first one, were you able to reformat and create a new partition and then check the health of the drive? did it work correctly?

Click to expand...

Thanks for the comments and Jim too.

Nope. I have not been able to do anything useful with either drive.

Most Windows tools are useless since they will not operate on a drive until it's been initialized. And Windows won't perform the initialization. DISKPART's "details" command crashes the program on both of these drives. Ick!

The other tools I've tried so far (e.g. Linux Mint, Gparted) don't see the drive and there's no /dev/xxxx/ to operate on there either.

I messed around with the WD utils on the WD drive and I couldn't get anywhere with that either. And since I have two other identical drives I was able to swap cables, and power bricks.

I'm now inclined to think the WD problem is of my own making. When messing with low level tools trying to diagnose/repair the failed Lacie, methinks I might have copied a bad MBR to the WD. It's my best theory so far. If so, very, very stupid mistake on my part, but at least I'm willing to admit to the possibility.

WD are not willing to offer any help beyond a RMA, so I'll probably do that. Later, when life is a little less hectic, I may disassemble the Lacie, extract the drive and plug it directly into a SATA port to see what happens. With ATA, we can see if a secure erase will clean things up.

I have run a couple more AV scans from standalone live disks and I'm getting fairly confident this is not a malware problem. Otherwise I'd secure erase my SSD system drive and clean install.

Yes, this is Windows 7 and I'm seeing identical behavior on my desktop and laptop machines so I don't think we have a BIOS or driver issue. And both systems are up-to-date on that stuff.

Richard said:

Richard wrote:

Trying to format the drives with another computer to format them.

Click to expand...

Done that. Same symptoms.

Richard said:

You said in previous post the drive was pulled out. Could be a USB problem. Try another computer and cable.

Click to expand...

Done that. On the WD I've tried another power brick too.

Richard said:

Do you have a magnetic field problem near your computer or drives, has the location of either changed recently.

Is your power surge protected, brown out protected? (good power)

Click to expand...

The location of the Lacie changed. Not the WD. The only close sources of electromagnetic radiation are my desktop, laptop, and a good quality UPS with surge/spike protection.

Member said:

malch wrote:

Click to expand...

[snip]

Member said:

The other tools I've tried so far (e.g. Linux Mint, Gparted) don't see the drive and there's no /dev/xxxx/ to operate on there either.

I messed around with the WD utils on the WD drive and I couldn't get anywhere with that either. And since I have two other identical drives I was able to swap cables, and power bricks.

I'm now inclined to think the WD problem is of my own making. When messing with low level tools trying to diagnose/repair the failed Lacie, methinks I might have copied a bad MBR to the WD. It's my best theory so far. If so, very, very stupid mistake on my part, but at least I'm willing to admit to the possibility.

WD are not willing to offer any help beyond a RMA, so I'll probably do that. Later, when life is a little less hectic, I may disassemble the Lacie, extract the drive and plug it directly into a SATA port to see what happens. With ATA, we can see if a secure erase will clean things up.

Click to expand...

Yea...

I'd remove them from their enclosures and plug them directly into drive ports, as sometimes the USB to SATA (or IDE) bridge chipsets can interfere with direct to drive functions.

For example, to even see SMART data, you may need to use parameters specific to the USB to SATA bridge chipsets (which can vary by the chipset an enclosure uses).

See this post (and the one following it) for examples of that kind of behavior:

http://www.dpreview.com/forums/post/50727732

So, I'd just remove them from their enclosures, plug them directly into drive ports from your PC's motherboard and see if you can see them that way. If it's just a problem with the partition table, you can use testdisk to recreate it.

I've accidentally repartitioned the wrong drive before when "half asleep" and managed to recover from that problem using testdisk (available in the software repos of most linux distros). Basically, it can scan a drive and try to determine the start and stop offsets of partitions and what type they are, then recreate a new partition table for you.

It can be a bit tricky on drives that have repartitioned multiple times with resizing of partitions (as it may find what it thinks are overlapping partitions due to past writes to those areas, where you have to make an educated case as to the correct offsets to use -- sometimes using trial and error to find the correct layout).

But, testdisk can be a life saver when a partition table becomes "trashed" for any reason.

I'd remove them from their enclosures and plug them directly into ports from your MB first if the drives are not showing up correctly as devices using something like fdisk -l as root (or sudo fdisk -l using an ubuntu based distro), as the USB to SATA Bridge Chipset may be interfering.

--
JimC
------

Member said:

Jim Cockfield wrote:
Yea...
I'd remove them from their enclosures and plug them directly into drive ports, as sometimes the USB to SATA (or IDE) bridge chipsets can interfere with direct to drive functions.

Click to expand...

I will almost certainly do that with the Lacie. However, the WD is still in warranty and I think it will be easy for them to tell I've cracked open the case. So I think I'll just RMA that one and let them repair or replace. If I caused the problem, I don't feel so great about that. OTOH, I'm not certain it was my fault, they won't talk to me, nor do they offer the equivalent of a factory reset button or utility.

I've bought a ton of their drives over the past 10 years so I'm not going to feel too guilty over it

My WD My Book drive has an eSATA port. Have you you tried connecting with that port, if your drives have one?

digitalshooter said:

digitalshooter wrote:

Testdisk help

Click to expand...

Yes, I am familiar with testdisk. In fact, it's entirely possible I used it to copy a bad boot sector from the broken Lacie to the good WD :-(

Next time, I will fire up one of my old machines in the garage for such tasks and not perform such risky procedures on a system with real data mounted!

I shall be eating crow and humble pie for dinner. Mmmmmm!

Pages 173 on Troubleshooting

If you are using a memory card reader that reads several types of cards you might check Windows Disk Management to see if you external drive is trying to use one of the "removable" drives that the card reader has reserved.

I had this problem with an external drive and solved the problem by changing it's drive letter.

The usual vilains for the problems you descibed are power source, cables and software drivers. It looks it is not your case.

Personally I gave up to use these external disk boxes and I went for a more direct approach: I bought a pair of USB3 dock stations compatible with HDD with SATA 1, 2 and 3 interfaces. I started to use just bare HDDs instead of full enclosures and I gained in reliability as all electronics and software usually will be simplier. Actually I dismantled my two WD Mybook and I am using their HDDs only and trashed the enclosures.

I am feeling much more tranquil with my approach although I had just a pair of failures of WD external disks in the past ten years - and for both cases I had the luck to have NO problems in the disks, only in the enclusure electronics...

Bottom line: perhaps you can solve your problems with a simple external USB3 dockstation as myself.

Best regards,

Osvaldo Cristo said:

Osvaldo Cristo wrote:

The usual vilains for the problems you descibed are power source, cables and software drivers. It looks it is not your case.

Personally I gave up to use these external disk boxes and I went for a more direct approach: I bought a pair of USB3 dock stations compatible with HDD with SATA 1, 2 and 3 interfaces. I started to use just bare HDDs instead of full enclosures and

Click to expand...

+1

I just use bare drives in a docking station. Bare drives are very easy to store (versus larger external drives with an enclosure, power brick, etc.)

Also, when I have a drive issue of any kind with an internal drive (for example, sector read errors indicating it's no longer reliable for day to day use, or SMART data telling me it's starting to have a problem), I just make a fresh sector by sector copy of the drive (cloning it, writing it's contents to a new replacement drive) and use the old drive as an additional backup.

That way, I can just plug the old drive into a docking station and access the files on it in the same state they were in at the time I replaced the drive if the need ever arises.

Docking stations are very handy for that kind of thing.

In fact, I'm making a set of backups (full disk image + separate home folder backup) from my wife's laptop that way right now as I'm typing this (using a mounted partition on a bare SATA drive in a USB docking station as the target for the backups).

My wife has a niece that's visiting with us, and she's been using the laptop. So, I'm going to make more frequent backups of it in case she messes anything up.

Even though I've got my wife's niece setup with her own username and folders (running Linux on the laptop) to isolate her program settings, folders, etc. from my wife's, you can't be too careful. So, I'm going to make more frequent backups while my wife's niece is here.

--
JimC
------

Member said:

Jim Cockfield wrote:
I just use bare drives in a docking station. Bare drives are very easy to store (versus larger external drives with an enclosure, power brick, etc.)

Click to expand...

Yes, at one point, I was all set to go this route for me primary backup needs. In the end I was dissuaded by the potential for problems arising with the SATA connectors which simply aren't rated for continuous removals/insertions.

However, I have also been noting the fact that quite a few people are able to make this work without problems. It looks like those connectors do have a decent life if handled carefully.

Obviously, there's still a finite risk of problems arising but I am starting to think they are small and outweighed by the considerable benefits of this approach. Time to buy a decent USB3 dock.

As an aside, bare drives do avoid some significant issues with the firmware that manufacturers (WD in particular) toss into their external drives.

I have been completely unable to find any malware on this machine which could have played a role in this saga. But I still didn't feel 100% certain the machine was clean.

So just to be safe, I have just secure erased my system drive (Samsung 840Pro SSD).

Windows has just finished installing so now I have a couple days of work to put all of my applications back and tweak everything to my liking :-(

OTOH, you just can't beat a nice clean system drive!

Member said:

malch wrote:

Member said:

Jim Cockfield wrote:

I just use bare drives in a docking station. Bare drives are very easy to store (versus larger external drives with an enclosure, power brick, etc.)

Click to expand...

Yes, at one point, I was all set to go this route for me primary backup needs. In the end I was dissuaded by the potential for problems arising with the SATA connectors which simply aren't rated for continuous removals/insertions.

Click to expand...

Look at the connector construction on popular Seagate and WD Drive models.

They appear to be very sturdy to me (touching the connectors, seeing how flexible they are, how thick/strong the plastic part appears to be, etc.). Perhaps the conductor portion (copper or gold etching) may wear out over time. But, the connectors themselves appear to be very sturdy.

The sturdiest connectors I see on my bare drives are newer Seagate models. The weakest connector I see (more flex, thinner construction) is on an older 320GB Seagate Drive, with the WD Drives somewhere in between.

But, even the drive with the connector that appears to be the worst of the bunch is still very strong from what I can tell. Given modern polymers (versus older plastics that are more brittle), I think you'd have to deliberately try to break a connector on any of them.

IOW, I just don't see one of them breaking by accident with typical usage patterns.

Now., I am careful with the drives (making sure I've got the drives turned in the right direction so that the connectors match up between the drive and docking station); and I also use the eject button on the docking station when removing a drive, while pulling on the drive at the same time I'm depressing the eject button (to reduce any stress on the connectors).

But, I think you'd be far more likely to damage the female socket or eject mechanism in a docking station, versus breaking the connectors on the drives, as from what I can see on the WD and Seagate Drives I have on my shelves in the office, they're very strong and well made, and very unlikely to break during typical usage patterns.

--
JimC
------

Member said:

Jim Cockfield wrote:

Look at the connector construction on popular Seagate and WD Drive models.

They appear to be very sturdy to me (touching the connectors, seeing how flexible they are, how thick/strong the plastic part appears to be, etc.). Perhaps the conductor portion (copper or gold etching) may wear out over time. But, the connectors themselves appear to be very sturdy.

Click to expand...

I don't disagree although if you check the specs on the internal SATA connectors they're rated for a pretty low number of cycles. I've seen numbers as low as 50 quoted.

But as a practical matter, they seem workable although it's obviously prudent to use great care when inserting or removing a drive from the dock.

Member said:

malch wrote:

Member said:

Jim Cockfield wrote:

Look at the connector construction on popular Seagate and WD Drive models.

They appear to be very sturdy to me (touching the connectors, seeing how flexible they are, how thick/strong the plastic part appears to be, etc.). Perhaps the conductor portion (copper or gold etching) may wear out over time. But, the connectors themselves appear to be very sturdy.

Click to expand...

I don't disagree although if you check the specs on the internal SATA connectors they're rated for a pretty low number of cycles. I've seen numbers as low as 50 quoted.

Click to expand...

Yea... But, I think most of the stuff you read about insertion cycles is referring to the cable connector design, not the drive connector design.

For example, scroll down the the eSATA section on this WD Interface Guide page, and you'll see this:

"eSATA and internal SATA cables and connectors cannot be used interchangeably. This is important since eSATA cables and connectors are designed for 5000 insertion and removal cycles while internal SATA cables and connectors are designed for only 50 insertion and removal cycles."

http://www.wdc.com/en/products/resources/drivecompatibility/

That's very vague; and I haven't seen anything in the actual hard drive specs about insertion/removal cycles for the drive connectors (versus cable connectors).

But, if you dig around, you can find more ambitious designs quoting far more insertion/removal cycles (when the SATA cables are taken out of the equation). For example, you'll find this on some enclosure manufacturers' pages:

"Reliable Up to fifty thousand times, the highest insert/removal cycle ever achieve in SATA connection - Guarantee FIFTY THOUSAND (50,000) insert/remove cycles, where ordinary SATA is only ten thousand times. Five time higher than any SATA backplane in the market"

http://www.cooldrives.com/mr4drqu3mosy.html

That enclosure uses a "tray less" design, meaning you're mating the drive's SATA connectors directly with the female connectors in the enclosure back plan), versus mounting the drive in a tray with separate connectors.

Of course, 50,000 insertion/removal cycles may be a bit optimistic. But, I wouldn't doubt the 10,000 figure for the drive connector itself (provided the etching itself holds up), if modern polymers are being used for the physical connector.

Again, just examine the drive connectors yourself. They look (and feel) to be very sturdy to me; and I've yet to have any issues using bare SATA drives with a docking station.

I am careful with the drives (making sure I don't have them turned around backwards, using the eject button while pulling on a drive while removing one to reduce strain on the components, etc.).

But, from what I can tell from the design, you'd be far more likely to damage something like the eject mechanism in a cheap docking station than you would to damage the SATA connectors on the drives.

As mentioned earlier, I think you'd probably have to deliberately damage a SATA connector on a drive (at least the WD and Seagate Drives I have). For example, take a pair of pliers, grab the end of the connector and bend it until it breaks.

IOW, I doubt you'd damage one by accident during normal use, even if you insert and remove one from a docking station multiple times per day over the life of the drive.

Damage to the female connector or eject mechanism in the docking station?

Perhaps, depending on it's construction quality and how you're using it. But, I suspect the actual drive connector is going to hold up over the life of the drive with popular WD and Seagate models, when they're being plugged in and out of docking stations (or SATA back planes in hot swap cages/enclosures, etc.); unless you're extremely careless (e.g., plugging a drive in backwards without the connectors mating together using a *lot* of force, and/or deliberately try to damage it).

--
JimC
------