W510 and USB 3.0 data corruption

EDIT:
I tried to edit the title of the thread, but it appears I can't.

I've determined that this is NOT Lenovo specific. I'm also having the problem on a Windows PC with a USB 3.0 add on card.

Has anyone ever used the USB 3.0 ports much on these W510's? I'm having huge corruption issues with them.

I was trying to use an external USB 3.0 drive to store a bunch of ISO images (hundreds of gigs of ISO images, mostly Microsoft stuff... OS's, server products, etc). Anyway, I have SHA1 files for each and use a program called HashCheck to verify the SHA1 of each file after I download it (Microsoft provides MD5's and SHA1's for each of their downloads on their msdn and technet sites). These are all downloaded and already saved to my NAS and verified good.

Anyway, lets say I copy 50GB of ISO's to the external hard drive, 9 times out of 10 at least one of the files does not match the SHA1. I can delete that 50GB and try again and this time a different file or files will be corrupt.

Here are the things I've tried. Nothing helps

Copied from network, corrupt.
Copied to local hard drive first, checked sha1's. All pass. Copy from local hard drive to USB 3.0 connected hard drive, corrupt.
Tried different hard drives, Toshiba 320GB 7200 RPM, Western Digital 160GB 7200 RPM, Seagate GoFlex 1TB 5400RPM. All fail.
Tried different enclosures (Vantec 2.5" USB 3.0, SIIG 2.5" SuperSpeed, GoFlex USB 3.0 cable). All fail.
Tried different USB 3.0 cables. All fail.
Tried two different W510's. I have one personally and liked it so much I made work buy me one for work use. Both fail.
One of my W510's still has the old 3021 USB firmware, the other has been updated to 3025. Both fail.

I've used the exact same USB 3.0 SIIG SuperSpeed enclosure and hooked up just a micro USB cable and plugged it into the USB 3.0 port on the notebook, effectively making it run at USB 2.0 speeds, and all works fine.

I've also taken the USB 3.0 cable/enclosure and plugged it into a USB 2.0 port on the notebook, and all works fine.

The same drive in an eSATA enclosure works fine (tried the 320GB Toshiba in a Vantec eSATA enclosure). Also got the eSATA cable for the 1TB Seagate GoFlex hard drive, eSATA for that one works fine.

I'm about to give up on USB 3.0 on this notebook completely. Unfortunately it was a part of the selling point, being a bit future proof by having USB 3.0 already.

What I don't have is another type of PC with USB 3.0 on it. I have a desktop but I'm not sure I want to waste even more money on this USB 3.0 technology and buy a USB 3.0 PCIE card if USB 3.0 is just not ready.

So, has anyone else done real testing using the USB 3.0 drives? If you haven't, I would question the validity of the data on your USB 3.0 drives until you do test it. I'm curious if anyone else can do some tests like this and see if they have any issues.

 

I copied 230GB the other day from a Seagate 500GB 7200rpm drive in the W510 ultrabay hard drive adaptor, to the same SIIG enclosure with a Hitachi 500GB 7200rpm drive.

No issues.

Sorry man, I am not seeing that issue.

 

Did you validate your data in any way? I can copy data all day long. Windows never errors out. But some of the data is garbage. What kind of data did you copy? Make some md5 or sha1 files to validate the integrity of the data. Create the files before copying then use those files to validate the integrity of the copied data.

 

All sorts of data. My docs, video, music, 100+ gig in .ISO files from my TechNet subscription, virtual machines, you name it.

I haven't seen anything come up corrupt.

 

It's hard to tell if something is corrupt. You may run into issues later. My advice would be to log into your technet account and note the MD5 or SHA1 for each of your isos, and then check them with something, such as HashCheck.

 

Looks like problem with your enclosure tobyg.
Personally I've experienced quite a lot of bad feedback about SIIG USB 3.0 hub (with different hosts, like PCIe cards), maybe it's their chips or SuperSpeed firmware are quite flakey. Or just niggles of whole USB 3.0 adoption being in quite raw state.

 

I've tried 2 enclosures, and also the Seagate GoFlex with the USB 3.0 cable.

I just bought a USB 3.0 card for my PC and encountered the same issue.

Not sure what the deal is, but it doesn't appear to be W510 specifically. So either it's all of my disks/enclosures, something else, or I'm the first one to find a corruption issue with USB 3.0. I find it highly unlikely I'm the first one to find an issue with USB 3.0, so it has to be something else.

Still, I'd like to know if anyone else has tested their USB 3.0 drives/enclosures to the level I have.

I just copied 300GB via eSATA, using the same 1TB GoFlex drive but with the eSATA cable instead of the USB 3.0 cable and all was fine. It's not the drive, it's not the system, it's not the source. I only have one conclusion.

 

Personally I think USB 3.0 is more a gimmick in the current time than something useful. That's why I've opted for eSATAp enclosure instead, lot less headache.

 

The thing that always annoyed me about eSATA was the need to have two cables, eSATA for data and USB for power. But now that Seagate (and Lacie has a model) have drives that have eSATAp cables and only need one cable, I'm going back to looking at using eSATA. The appeal for USB 3 was the need for only one cable.

Meanwhile, I found there are newer NEC (now Renesas I guess) drivers. The latest I found was 2.0.4.0. Testing these now. I was running the stock Lenovo ones, 1.0.18.0.

 

You should use the extra power cable with the SIIG JU-SA0312-S1 as well.

 

Sorry, I forgot to mention that. I did try the extra power cable.

I actually have two SIIG cases (tried both, bought them a long time ago but never used them this much/noticed the problem, so I can't return these), had a Vantec case (returned it as it was causing me the same issues) and currently still have the 1TB Seagate GoFlex drive with the USB 3.0 cable (although I'll probably return the USB 3.0 cable, I'm giving up on USB 3.0). The 1TB Seagate GoFlex drive I'll probably keep, and use with the eSATAp cable. It works great that way.

The 2.0.4.0 drivers I found also did not help.

I'm starting to think there is a reason Intel still hasn't put USB 3.0 on their motherboards. And that's probably also why Apple hasn't gone with USB 3.0 yet. They're usually on top of new things such as USB technologies. The only new thing Apple seems to be not going with, besides USB 3.0, is Blu-Ray. And Blu-Ray isn't very new anymore.

I don't understand how it can be anything other than a problem with USB 3.0. Using the same drive, same computer, same source material, same everything minus the cable (in the case of the Seagate GoFlex drive) and when I use the USB 3.0 cable I get corruption, but using eSATAp I don't.

It's either a USB 3.0 hardware problem, in the USB 3.0 chips in the computers, or in the drive/enclosure, or a problem with the drivers.

But I'd like to caution everyone into making sure they validate their data. Just because Windows copies it, just because there are no errors, just because everything looks right, it may not be. For my test I downloaded this 1.7GB zip file (EMC Celerra VMware image) and had 20 copies of it on my NAS device. Created an SHA1 for all 20 (I just copied it to my NAS as 1.zip, 2.zip, 3.zip, etc) and then when I copy it back I tested them with HashCheck. Some of the zips fail. Zips are good because you can then also use some zip program (I use 7-zip) to 'test' the archive, and it will fail also.

I encourage you to all test your USB 3.0 drives/enclosures this way.

Your data may have corruption, and you may not even realize it or want to admit it.

 

I ran SHA1 calcs against Win7 Ent, Pro and Ult x64 .iso's. I ran calcs against Windows Server 2008 R2, SQL Server 2008 R2, and Office 2010.

All of mine are checking out fine on the Seagate drive in my SIIG case, and the Hitachi drive in the ultrabay which was originally in the SIIG case.

No corruption here.

 

Well just to butt in, Apple really isn't on top of new things all that often (lately). They haven't done BluRay, Wireless USB, USB 3.0, eSATA, 3G, Wimax, and they were slow to do SD readers, Displayport, higher resolution screens, and other stuff. And they aren't some sort of God when it comes to standards either, Firewire 800 comes to mind. There's almost nothing out there that even uses Firewire 800 exclusively (maybe the $40,000 Hassalblad camera )

I guess your USB 3.0 setup isn't working so great, but other than the Hash check, do you have any other evidence? Like try opening them or using them.

 

A hash check is enough evidence for me. If the file isn't identical to the file I copied, why would I dare use it? The problem is, a hash check is only good for static data. I was using VM's and trying to do Acronis backups to my USB 3.0 drives. Every time you use a VM, the VM files will change. And I no longer trust any backups I made to these USB 3.0 drives.

I think my problem may be heat related. I'm still testing but if I keep a fan blowing on the open enclosure, things are better.

This may be related:
SuperSpeed USB 3.0 - Problems & Issues with USB 3.0 SuperSpeed | Premium USB Blog

But they say "the chips are performing beautifully despite running a little too hot." Not sure about that. I wonder if they truly are running fine.

 

Yeah, I'm pretty sure this is heat related.

I was doing more tests tonight and all was working pretty well. I wasn't seeing errors. I was testing to a 320GB Toshiba drive in the SIIG enclosure.

Then I tossed the 1TB Seagate drive on top of the SIIG external chassis and that kept the SIIG a bit warm (not burning hot, but definitely on the warm side. The heat that comes out of the side of the W510 is still warmer (And maybe this is some of my problem, I usually have the drive near that side, but I don't tonight).

Anyway, when I did this, failures happened all over the place.

Here is an example.

These are those 1.7GB files I mentioned. They are all the same. You can see though, many of them are corrupt, all with different CRC's.

I might also add, I've never had 'read' corruption. Only write. Once a file gets written, I can read it back fine. And by fine I mean, if I retest these, they will all either pass or fail the same way, same CRC's as the first test. I hope that makes sense.

Attached are some pictures to show what I see.