Strange External Drive Problem

[pmokover]

I’m having a strange problem with an external hard drive. I have two of the same model drive and one of them has a problem. I will refer to them as drive #1 (good) and drive #2 (problem). The drives are 1TB, USB3. My PCs are running Windows 7 and they all have USB 3.0 ports.

I have one of these drives connected to my main PC at all times but only one of them at a time - never both at the same time.

Drive #1 always works perfectly.

When I try to write to drive #2, 90% of the time it “stalls” for as much as a minute before it starts writing. (10% of the time it works normally.) This happens whether I manually copy a file to it using Windows Explorer or use a program to save files to it. It just stalls but most of the time it eventually un-stalls and writes.

This is what I have tried so far:

1. Tried drive #2 on two other PCs. Same problem with both other PCs.

2. Tried drive #2 with different cables and different USB3 ports. Same problem with all.

3. Checked Windows 7 security settings. All OK and all are set the same for both drives. I also checked the various properties settings for the drives and everything looks correct there. I also checked for an updated driver and I already have the latest. I also have all of the latest updates for Windows 7.

4. Tried drive #2 (which is USB3) on a USB2 port. Same problem.

5. Re-formatted drive #2 and started from scratch. The problem remains.

6. Contacted the manufacturer (LaCie) and tried everything they suggested. The problem remains and they have no additional clues.

7. I returned drive #2 and ordered another one of the same model from a different vendor. The replacement has the same problem.

8. I returned the replacement and ordered another replacement from the manufacturer (LaCie). The second replacement has the same problem.

All this time drive #1 (same model) continues to work perfectly.

I’m baffled by this one. Any clues appreciated.

Thanks.
Peter

[falther]

Sounds very strange and I see only two points left to prove whether its a LaCie or a software
issue:

1.) Connect a SATA hdd with a USB 3.0 SATA to USB 3.0 adapter - or use any kind of SATA to
USB 3.0 enclosure for this hdd - and connect that to your pc´s.

If that works perfectly, then you could try the following procedure with the LaCie hdd:


2.) Low level the hdd and create new partition(s). Does LaCie offer any kind of software
tool to prepare the hdd (partition, test, low level)? If not - you could do it in Control Panel and
in (sorry dont know the exact translation) System Management (?) - and there you should find
a kind of Drive Management (or similar expression) to create or modify or format partitions etc.

Low level is the most important part here. If you did that allready with LaCie tools without
positive result, use the low level only - and then continue with the Windows tools to partition
and format - avoid existing LaCie tools.

+++

[pmokover]

Falther,

Thanks for the suggestions.

I don't have a USB3-SATA adapter but I will get one and try that.

I did the low level format but that does not seem to have helped.

Peter

[Vulcan]

@pmokover: Interesting problem. Have you checked the Windows system event log, for any messages which are logged during that (up to one minute) "stall"?

[pmokover]

An update:

I ordered a different brand/model 1TB, USB3.0, external hard drive (Western Digital "My Passport Essential") and it works perfectly. So, in summary...

1. My first LaCie Rugged 1TB, USB3.0 drive works perfectly.

2. Three others of the same model LaCie drive (purchased from three different vendors) have the problem described in my original message.

3. My WD 1TB, USB3.0 drive works perfectly.

Looks to me like LaCie may have a bad batch of these drives.

Peter

[Vulcan]

@pmokover - Interesting info, thanks. Just one question - have you ever tried attaching one of the "problem" disks to a PC, where that PC has not had the "good" disk attached to it first? My hypothesis would still be a LaCie problem, but it would not be a disk problem, and would not affect people who have only 1 of those disks attached to any given PC.

[pmokover]

@pmokover - Interesting info, thanks. Just one question - have you ever tried attaching one of the "problem" disks to a PC, where that PC has not had the "good" disk attached to it first?

I *think* one of the PCs did not have the good drive attached to it first but I am not positive about that.

Peter

[pmokover]

Connect a SATA hdd with a USB 3.0 SATA to USB 3.0 adapter - or use any kind of SATA to
USB 3.0 enclosure for this hdd - and connect that to your pc´s.

I finally got a USB3-SATA adapter and connected an old drive and it seems to be working OK.

Peter

[Vulcan]

I finally got a USB3-SATA adapter and connected an old drive and it seems to be working OK.

Very interesting, Peter. That's just what I was expecting when I wrote my last comment - i.e. not a disk issue. Congratulations on finding that result, using falther's excellent suggestion.

[pmokover]

So are we concluding from all of this that LaCie has a bad batch of this model drive? All of the evidence points to that but its still hard to believe that three drives from three different vendors all have the same problem while one other drive of the same model does not.

Peter

[Vulcan]

So are we concluding from all of this that LaCie has a bad batch of this model drive?

That depends what you mean by "this model drive" - do you mean "disk only", or do you mean "disk and enclosure"?

I thought your previous comment had confirmed that the disk (from a previous "problem" enclosure) was actually OK, when you used it with a different USB-SATA converter (i.e. not the one from its original LaCie enclosure), as falther kindly suggested.

[pmokover]

I thought your previous comment had confirmed that the disk (from a previous "problem" enclosure) was actually OK, when you used it with a different USB-SATA converter (i.e. not the one from its original LaCie enclosure), as falther kindly suggested.

A little misunderstanding here. The problem LaCie drive is a self contained sealed unit and I have tried it only that way. I never tried the drive inside with a different adapter because I did not want to break the seal (which would kill my option to return it).

The drive I tried with my new USB3-SATA adapter is a hard drive I no longer use from one of my desktop PCs. I thought the purpose of that test was to see if a different drive would work with a different USB3 adapter (which it did).

I don't have any way of knowing whether the problem with the LaCie drives is the drive itself (Toshiba) or the internal USB3 adapter. In my previous messages when I referred to the LaCie drive I was referring to the sealed package of the drive+adapter as it comes from LaCie.

Peter

[Vulcan]

Hi Peter, Thanks for the clarification - as you said:

A little misunderstanding here.

Indeed, I misunderstood. When you said:

I finally got a USB3-SATA adapter and connected an old drive and it seems to be working OK.

... I had interpreted that "old drive" to be one of the "old" LaCie drives (which you must therefore have taken out of its case, since you were using it with a different USB-SATA bridge).

The picture is a little clearer now, but I doubt that I'll get to root cause without having the drives here. There are many possible troubleshooting approaches I can see, depending on available time, money & equipment. I'll suggest some thoughts that I have below, but as I said, the limitations of remote diagnosis along with the lack of info about the USB or SATA I/O during the "stall", are a problem IMHO.

As I suggested before, looking at what the system is doing during the (up to 1min) "stall" when using the "problem" drives, would quickly point in the direction of the problem - i.e. what is actually causing that "stall". I realise that another piece of unclear info is whether that "stall" occurs for every access to a "problem" drive; or just the first access after the "problem" drive is attached; or just the first access to a "problem" drive which needs the drive to spin-up; or something else. You mentioned that the "stall" occurs with writes - what about reads? Does the "stall" occur when a "problem" drive is the first one attached to a PC after a power-on, and not following from the "good" drive being attached? The timing of when you're attaching "good" and "problem" drives during the testing, wasn't clear to me from your earlier comments. All these answers may also tell you something, about when the "stall" is, and is not, triggered.

Techniques for examining what is happening during the "stall", would include using a USB or SATA protocol analyser - but hardware protocol analysers are expensive; software-based USB protocol analysers exist, but then you've still got to spend time looking at what gets captured, and interpreting it. Earlier I also suggested looking at the Windows system event log, to see if that gave you any clues.

Another piece of unclear info, is whether this is a USB-powered drive, or an externally-powered drive - I saw LaCie 1TB USB3 drives of both types advertised. Although USB3 increased the power which can be drawn from a port, if yours is a USB-powered drive, could a change in power requirements between the "good" and "problem" drives, be involved in the behaviour that you see?

As I understand the results of your recent test, due to the limitation that you don't want to open the LaCie enclosure, the only conclusive test result is that you don't have a generic USB3-related issue on your PCs, due to the different USB3-SATA bridge working OK; but we knew that already, since your "good" USB3 drive works, didn't we?

I was wondering whether the other slightly unusual part of your config, could be related i.e. you're connecting multiple identical drives (individually) to the same system. Most buyers of these drives won't be doing that, will they? Therefore I'm thinking that you could be affected by something which other users (and also LaCie support) may be unaware of. Are these different drives being correctly recognised as different drives - are they reporting different USB h/w serial numbers? At one stage, when I had more Windows driver involvement than I do now, I think it was a requirement that all mass-storage devices did have unique USB serial numbers, if you were going to connect more than one to a system (or to swap between them).

What opportunities do you have for trying the "problem" drives (NOT the "good" one!) on other PCs, even just on a USB2 port? If you can find any other PC where the "stall" does not occur (doing whatever test causes that "stall" on your normal PCs), then that would be an important result.

So those are the types of areas I would investigate, if I was in your situation, with the drives in front of me (except I'd also be using a protocol analyser to look at what happens during the "stall"). Hope that gives you some ideas.

[pmokover]

Vulcan:

Thanks for the very detailed reply!

Before I address your individual points I should start by saying that the subject is now purely academic as I have returned all three bad LaCie drives. I still have the one good drive (plus a new Western Digital 1TB USB3 drive which works perfectly).

I did not have the chance to look through the system event log to see what was happening during the stall. Perhaps that would have shown a clue to this problem. I also did not try a USB protocol analyzer as I don’t have one.

This particular model LaCie drive is USB powered.

I tried the bad drives on several different PCs with the same result on all. I also tried them on USB 2 ports with the same result.

I did consider the possibility that the problem had something to do with the fact that I had connected the good drive to the PC first and later connected the bad drives (same model). I got the good drive about a month before I got the first of the bad drives.

The PC did correctly see the serial number of each drive and each drive had a different serial number.

I should point out that this is the only problem I have ever had with a USB drive. I have used several models including a few other LaCie models. No problems with any of the others.

As I said, it’s academic now but I’d still like to know what caused the problem. Thanks again for your help.

Peter

[Vulcan]

Hi Peter,

Thanks for the update and additional info.

As I said, it’s academic now but I’d still like to know what caused the problem.

Understood. Given that it's now impossible to gather more results (e.g. the tests around when the exactly the stall occured, that I mentioned) or to test any theories to see if the "problem" drives start to work OK, then I don't believe we can go any further.

The PC did correctly see the serial number of each drive and each drive had a different serial number.

Depending on where you saw that serial number, it is very very unlikely to be the one I was referring to. If you were using any kind of disk utility, especially one which also showed the disk make & model, then it was almost certainly the disk's own serial number that you were seeing. It's somewhat confusing that there are several different serial numbers...!

I was referring to the USB serial number, it's part of the USB-SATA bridge functionality. That is what is used by Windows to identify different USB-attached drives which have the same USB VID & PID (or at least it was when I looked, many years ago) - and I expect that all of your LaCie drives of the same make & model, had the same USB VID & PID; that would be typical. Hence the USB serial number needs to be different in each of those drives for Windows to distinguish them - but this is only important if a user has more than 1 of those drives. So that is where I had reached in my thinking, given that you're relatively unusual in having more than 1 of the same drive, and swapping them.

That USB serial number can be found in the Windows registry (it's recorded when a device is first attached), or can be read by a few obscure USB utilities. Now you see why I (mistakenly) thought your testing of an ex-LaCie drive, with a different USB-SATA bridge, was giving us a pointer in the direction that the original LaCie USB-SATA bridge was involved in the problem.

Of course if this theory were true, it would still be a LaCie problem, with the programming of their USB-SATA bridges, but it would only affect customers who had multiple drives of the same model and who swapped them while the PC was running. That may explain why the support folks you spoke to, hadn't heard of problems, due to the relatively small proportion of customers who would be affected. It may even be a batch problem with the programming of some USB-SATA bridges, so the affected number of customers would be even fewer!

It's only a theory, but it seems to fit some of the facts and hence was worth considering further, IMHO, if you still had the "problem" drives.

Anyway, at least you've got 2 working drives now. Good luck!