corrupted external usb drive - what went wrong?

[briligg]

A few days ago a folder on our external hard drive suddenly contained several mystery folders and files with nonsense names of odd characters, in total containing 6gb of unreadable stuff, which i couldn't delete as the file names were invalid. This drive gets swapped on occasion between my machine and my husband's machine, and although my win 7 machine asked me a few times if i wanted to scan and fix problems on the drive, i ignored it thinking there was just some issue about swapping drives between xp and 7 - it asked me that whenever the drive had been plugged into his computer.

I ran virus scans on both computers. One was removed from hubby's machine, mine was clean. I did a registry clean up with CCleaner, and started a free space wipe on the drive in question, but stopped it when i saw how much time it would take, to check into if that was a good use of time. At that point, i opened the corrupted folder and the bad files were gone.

Hm. okay. And windows wasn't asking me to scan and fix the drive anymore, so i thought everything was peachy. Then a bunch of win 7 updates were installed on my machine yesterday, and when i booted up afterwards it couldn't read the external drive anymore. Nor could hubby's machine. So, i backed up some key stuff onto a thumb drive - and for the first time, when i plugged it in, i was asked to scan and fix the drive... which this time i did. That drive has also recently been plugged into hubby's machine.

Now i've formatted the external drive, and i'm putting it through a wipe with CCleaner. I just plugged in another thumb drive which has never been on the other machine, and it opened fine - but on a different usb slot, the other one is busy until the drive wipe is done.

So, what is the issue here? Could it have been the virus on hubby's computer? A hardware issue? What should i do next?

 

[Emulex]

i find using quality SHORT cables and not overloading the usb bus to help alot. alot of the controllers are powered by the usb (odd since you have a power cord) so if you droop the power by plugging in too much - corruption.

also vibrations. i had a guy put a hammer down about 1ft from the drive while active and it died. not slammed but enough to make me winch and GRRRRRR

 

[Echo147]

Are you "safely removing" it from the tray after use? Not so important if just reading from the drive during that session, but writes are likely to cause corruption if you just pull it out, even a few minutes after activity.

 

[briligg]

No, i always unplugged it after closing explorer, when i couldn't see how there would be any writing or reading activity. The corruption appeared in the folder of a movie we had played a day or two before i noticed it. But i don't think i unplugged it after watching the movie. The drive definitely could have received a good bump where it sits under my desk, that could have happened while it was doing a read or write.

But the thing that really concerns me is i thought the problem was fixed. The corrupted section had been deleted, and my machine seemed to be able to read it fine. (How the file system was suddenly corrected, i don't know. Maybe during the few minutes CCleaner wipe ran, it covered that portion?) Then after the Windows updates, suddenly it couldn't see anything but an unformatted drive.

I now take heed of the safe eject button... It still doesn't seem like that could have caused an issue of this magnitude, especially since there was no unplug/replug before the whole thing suddenly failed.

It usually is plugged into a short extension cable running to one of my 2 usb 3.0 slots. The usb 2.0 slots on hubby's machine have failed before, but that is supposed to be fixed, and he hasn't had any problems with them. Only 4 of the 10 slots on my machine are being used, so i don't think the bus is overloaded.

 

[briligg]

It now seems to be working fine, i have started reloading all the things that were backed up onto it. But can i trust it?

 

[briligg]

Maybe i actually should have posted this in the computer help section. I'm getting the feeling now that the average tech level in this forum is way higher than mine.

 

[Cerb]

I now take heed of the safe eject button... It still doesn't seem like that could have caused an issue of this magnitude, especially since there was no unplug/replug before the whole thing suddenly failed.

Oh, it could have, it just took awhile to manifest. Say no to checking the drive only if you have very good reason to (that data loss could be catastrophic, because you don't have other copies, and you know Linux geeks). That prompt is coming up because Windows had reason to believe that there may have been errors in the file system.

 

[Emulex]

Oh, it could have, it just took awhile to manifest. Say no to checking the drive only if you have very good reason to (that data loss could be catastrophic, because you don't have other copies, and you know Linux geeks). That prompt is coming up because Windows had reason to believe that there may have been errors in the file system.

NTFS policy write cache = high likelihood of corruption esp due to bad cable, must eject, faster performance non-linear

fat32/quick eject policy - immediately write and sync (no lazy cache), not so dangerous, but you should still eject.

also keep in mind when the drive is ejected it may (or not) spin down. this takes time. don't move the drive around until it has settled.

I used 2TB freeagent's for D2D2D backups at work and they have been reliable AFTER burn-in - but two of them failed within 1 cycle (full format not quick, full write all 2tb). the other 7 did not other than the hammer problem.

keep in mind BASS from a subwoofer = very very bad. these are low quality drives (external cased drives are the lowest quality bin). They are high areal density. that spells easy disaster.

And heed my warning on using a quality shield SHORT usb cable and don't put anything else on the same usb controller. (usually one controller handles 3 to 4 ports). trust me here. don't plug in keyboard/mouse/drive to the same controller. the key/mouse could hang the bus and cause corruption or suck down the power that SOME usb controllers use.
Shielded cable also prevents noise. Cheap usb cables (IE freeagent white cable) is unreliable imo. I use a 9" cable that is heavily shielded, thick, good tight connections on all the freeagents now. i had numerous problems before.

If you are serious i'd suggest using a pci-express x1 usb card. so it has full power and is totally isolated from the other devices. this would cost very little.

or go firewire - it is about 10x less fragile (still have to eject). plus goes to 800meg.

esata works too but you have to eject and its even more sensitive i limit my esata cable to 6" max. unless you are using infiniband quality cable like SF8088 - then you can go several feet. SF8088 quality cables can be $70-100 for really good quality.

 

[lsv]

OP I'd run some diagnostics on the drive. It could be cable it could be the drive itself or even the USB controller. Do you have any way to test the drive internally over IDE or SATA?

 

[Cerb]

NTFS policy write cache = high likelihood of corruption esp due to bad cable, must eject, faster performance non-linear

fat32/quick eject policy - immediately write and sync (no lazy cache), not so dangerous, but you should still eject.

FAT32 = reformat NTFS, IMO. NTFS doesn't wait long to write to external drives, but Windows doesn't give a good indicator of when it is safe, yet it could be writing a little here and there, many seconds after anything on-screen is gone.

I used 2TB freeagent's for D2D2D backups at work and they have been reliable AFTER burn-in - but two of them failed within 1 cycle (full format not quick, full write all 2tb). the other 7 did not other than the hammer problem.

I assume any portable HDD is on its death bed, unless treated very carefully. Penny pinching has gotten the best of every major manufacturer, now.

And heed my warning on using a quality shield SHORT usb cable and don't put anything else on the same usb controller. (usually one controller handles 3 to 4 ports). trust me here. don't plug in keyboard/mouse/drive to the same controller. the key/mouse could hang the bus and cause corruption or suck down the power that SOME usb controllers use.

I only use my ports through 6' extensions, lately (past 2-3 years), for hot-plugging, with no problems. I have sliced open a few (it's not like I'll ever run out), and the differences on the inside tended to mimic the outside. I've never bothered trying to figure out what port goes to what controller. Just not using hubs (or putting all non-storage peripherals on a hub, when USB ports are scarce), and sticking largely to Intel and AMD chipsets, tends to prevent problems (740G marked the end of excuses for others' mainboard chipsets, IMO, though add-on cards could be another issue).

Shielded cable also prevents noise. Cheap usb cables (IE freeagent white cable) is unreliable imo. I use a 9" cable that is heavily shielded, thick, good tight connections on all the freeagents now. i had numerous problems before.

WD's external cables are just as bad, IME. I almost RMA'd a Elements once, and the problem turned out to be the included cable. I can only imagine the opinions of the engineers that don't get to spec and source the included cables. I wouldn't even know where to buy cables as thin and flexible as what come with those drives, nor that offer such flimsy connections so often.

esata works too but you have to eject and its even more sensitive i limit my esata cable to 6" max. unless you are using infiniband quality cable like SF8088 - then you can go several feet. SF8088 quality cables can be $70-100 for really good quality.

What makes eSATA so bad? I've been using eSATA daily for >3 years, now, with 4' and 2m cables, and not a single problem. At 6", the interface is pretty much worthless. It's not like it's not differential, lacks CRC checking, or that there is a short limit on the spec.

 

[Emulex]

esata cables are the bad. not the esata itself. a great esata cable (like a great usb cable) makes a world of difference. unfortunately people buy cheap cables and think they are all the same. they aren't. plus some esata connector ends are poorly made too.

you are right - penny pinching is what the manufacturers do these days and it causes problems. when you are pushing terabytes a day to drives you tend to shake out the ugly really fast and nobody wants a corrupted backup.

 

[briligg]

Emulex - i've removed the cable and plugged the drive directly into a usb 3 slot, and moved my phone from the 2nd one to another slot. So, the drive should be the only thing on that controller, right? Plugging and unplugging will now have to be done by feel while i crouch under the desk, but it's feasible, considering the drive is not swapped often. The drive is also now on top of the computer itself, where it is safer from bumps.
My motherboard is a Gigabyte X58A-UD3R, which boasts of 3x the power to the usb ports, for what it's worth. And the drive is Adata's ruggedized version, but i suppose that only helps when the drive is unplugged.

lsv - the drive is only usb. can you recommend a diagnostic tool?

 

[briligg]

So, i was doing a system backup, and i got the above error. There is still a corrupted, unreadable spot on the disk. I guess my abuse of it added up.

So far i haven't found anything i can do about it. Any advice?:\
The tips so far are, by the way, appreciated.

 

[C1]

Download and run the free SeaTools. Run the detailed/long scan/analysis. If a bad sector(s) is found then SeaTools will report these and provide an option for repair (ie, map around/out any bad sectors.

 

[briligg]

It passed a 'long generic' test in Sea Tools. Could the problem be something other than the drive?

 

[briligg]

Ok, and immediately after passing that test, windows now once again can't see anything on the drive, and is asking me if i want to format it.

 

[Emulex]

sounds like the bad cable/power supply issue i've seen.

swap the power supply and/or usb cable mate.

 

[VNGrunt66]

Being somewhat new at this, could we get a couple of suggestions on good USB cables? (i.e. brand names or particular part #)

Thanks

 

[C1]

I use "ZIO" with all kinds of drives and never had an issue.
https://www.andovercg.com/store/secure/images/products/zio-usb-cable-extender.jpg

 

[VNGrunt66]

I note that many USB cables refer to "high speed USB 2" in their description.

I just installed a USB 3 PCIe card and purchased a USB 3 HDD enclosure and wondered - will existing USB2 cables do the job well enough, or will their be a change for USB 3, besides just re-naming the cables?

Thanks

***** Went and did some Googling and found out that USB 3 cables are different from USB 2. Extra conductors within the cable used by USB 3 are necessary to use USB 3 devices and the cables are generally fatter due to this.

So, the answer to my inquiry is that one needs USB 3 cables to run USB 3 devices in that mode. ****

 

[Cerb]

Being somewhat new at this, could we get a couple of suggestions on good USB cables? (i.e. brand names or particular part #)

Thanks

Monoprice's that mention shielding, thick and sturdy dollar cables, etc.. Seriously, some external HDDs come with cables that have worse connectors than those I've bought at the dollar store. Just like with analog ICs, you get what you pay for, and the trick is to pay several dollars for a decent cable (or $1 discounted ), rather than the cheapest way-under-spec cable with expensive marketing. IME/IMO, Emulex might be being a bit paranoid, but it is annoying how bad some cables can be.

If it properly meets specifications, and the connectors are made of good materials, with decent plating (if you can accidentally bend parts of the connector, that's not a good sign, and that happened with one including WD USB cable once ), it aught to be good enough...but it's not like you have a lab at home with GHz-range signal testing equipment. It's really not rocket science. Coming up with the specifications, sure, but meeting them is more about copper, brass, and aluminum alloy costs, and manufacturers preferring to lower costs when their margins are thin, even at the expense of product reliability (a badly manufactured cable that happens to meet spec at the factory could fail after months of use, FI, thanks to subpar materials and/or treatment of those materials). Prior to a couple years ago (I'm sure it's very much a recession thing), both WD and Seagate had sturdier cases all-around, with good ventilation, didn't has nearly as much annoying crapware, and it wasn't until '09 or '10 that I started noticing any problems with the included cables from WD (haven't used any external Seagates since '09 or so; all been WD or Toshiba, as of late).

 

[briligg]

I am only using a USB 2 cable, which works but doesn't give me the extra speed. But at least it is on its own controller, i guess.

I did a check disc on drive C, after pondering that the drive was failing when windows tried to backup, but not when other files were copied to it. All it found was some unused index entries.
I found another cable that fits in our stash of assorted parts, it seems a little sturdier. I also moved the cable to the other USB 3 port.
Now for the first time since this began, windows has successfully backed up.

Where i am it isn't the easiest thing to buy cables. Nor am i eager to spring for that unless that is definitely the problem. The drive worked fine for a few months before this problem cropped up, despite me doing a bunch of bad things to it. Because of that, actually, i don't know if i should blame the cable or the port, or keep looking for another problem.

 

[briligg]

New weirdness - the computer has been on for hours, i haven't done anything with the external drive, and suddenly a message pops up that a driver has been installed for the mass storage device and it is ready to use. I went straight to check the files, they are still being read properly. I haven't been doing anything but surfing. I don't have any automatic update software for drivers. What the hell?