Momentus XT, Momentus, Momentus Thin, MobileMax, and other notebook drives

[Ironclaw]

There has actually been a (probable) reproduction on a Mac: http://superuser.com/questions/313447/seagate-momentus-xt-corrupting-files-linux/315359#315359

Also, in comments section of the question, there was a (probable) failure to reproduce on Linux.

Possibly related: http://forums.seagate.com/t5/Momentus-XT-Momentus-Momentus/More-XT-500-Data-Corruption-with-Linux/td...

---

gazoi wrote:
perhaps Seagate just swapped over the stickers and gave you the same drive back

---

Wait. Now that I think about it... The first drive that I sent back did not fail any traditional tests. I think it is possible that Seagate thought I was an idiot for sending back a seemingly flawless hard drive, and simply changed the original label to a refurbished one. I don't think it is likely though. And almost certainly not the case for the second drive I sent back, because I accidentally bricked that one while downgrading the firmware.

---

gazoi wrote:
Have you heard any more from Seagate tech on the matter?

---

I have not heard anything more.

I do think it may be a good idea for you to RMA the drive. If you get back a drive with no problems, then we've sort of solved the problem.

However... I've just made up my mind to submit something to Slashdot and Digg about this. I don't know if it'll be accepted on Slashdot, and it might not be noticed on Digg, but it's worth a try to get more people testing this. I will submit within 24 hours; you might want to wait and see if others are able to reproduce this.

[gazoi]

This is interesting. I just tried to reproduce the infamous file corruption on my day in day out system which is Fedora 14.

Linux elegia 2.6.35.13-92.fc14.x86_64 #1 SMP Sat May 21 17:26:25 UTC 2011 x86_64

x86_64 x86_64 GNU/Linux

This was using the Momentus XT as an external drive over USB2, copying a 4GB file.

Guess what, all good, I stopped after 20 iterations. All my other tests were using the drive internally. The problems were with Fedora 15.

I'm just downloading a Fedora 14 image so I can install it on the drive and test then.

I thought other people were getting file corruption when using the drive externally too? Or maybe this is really actually linked to certain Linux kernels?

[Ironclaw]

I think that's probably because of the USB connection. I've never been able to reproduce the problem when I connect through a USB 2.0 interface, and as far as I know, no one has had file corruption through a USB connection. I was mainly testing through an eSATA interface on my external drive enclosure.

It is possible that older kernels do not have the same problem, but so far I don't think we have any evidence pointing to that. I once tried testing with older releases of Ubuntu, but they didn't support my laptop.

I have submitted this to Slashdot, Digg, and Phoronix forums over two days ago, but there has been no response.

[frief]

> and as far as I know, no one has had file corruption through a USB connection.

My test (which reproduced the problem) in this thread (2011-07-12) was done with an USB stick.

[Ironclaw]

---

frief wrote:

My test (which reproduced the problem) in this thread (2011-07-12) was done with an USB stick.

---

Did you mean that the OS was boot from a USB stick? I was referring to tests done with the Momentus XT connected through a USB interface.

[frief]

No, good you ask back. OS and target drive was the Momentus XT, only the file with the pseudo-random data was on the stick.

[frief]

I feel somewhat ashamed for asking the self-evident:

Could it be acknowledged by SEAGATE, that fixing this easy to reproduce and well documented BUG is a prerequisite for releasing the next firmware revision?

[gazoi]

I have just received a new drive back from the suppliers after returning the old one a couple of weeks ago.

This was manufactured in 06/11 and has firmware SD26 pre-installed. The SMART data shows lots of previous errors on the drive and there are a few scratches, so it's definitely a reconditioned drive. I've always wondered why SMART data doesn't get reset when they do a factory reset.

Anyway, I just ran some tests externally over eSata, 25 x 4Gb files and I found no corruption. This is with Fedora kernel 2.6.35.13-92.fc14.x86_64.

It woud appear that either SD26 has fixed the problem, or maybe it was actually just a dodgy drive after all? I guess if you still have a problem, you need to get hold of firmware SD26 and see if that changes anything.

I'll test the drive internally next when I get a chance and post my results back.

[Ironclaw]

Thanks Gazoi. I have just flashed my firmware to a leaked copy of SD26, and I can confirm that it seems to have fixed the problem with file corruption. This was on an ext4 partition over eSATA on Ubuntu 11.04. The test was looped for 35 times with a 4GB file.

I think I will send an email to Seagate technical support to ask them what's going on. If they are aware that SD26 fixes a problem with file corruption on Linux, then it is unacceptable that they did not make a public announcement. Otherwise, if they are unaware that SD26 fixes a problem with file corruption on Linux, well... that is even worse.

I have found leaked copies of both the Windows update utility (.exe) and the bootable update CD (.iso) for SD26. I would be happy to provide these to anyone - just reply here or PM me. However, I would recommend asking Seagate for a copy of SD26 instead of getting a copy from me or anyone else, because you cannot verify that I have not tampered with the update. If you would still like to get the leaked copy, the following paragraph tells you what I have done to protect myself from malicious tampering; you might want to do the same.

The leaked Windows utility and bootable CD are from two different sources, and I have extracted their firmware images and verified that they are identical, so it is unlikely that the firmware contains a modified malicious payload. I have also checked the contents of the leaked bootable CD against the official SD25 bootable CD, and again I don't see any malicious tampering. If there is a malicious payload on the leaked CD, it has to be hidden within the metadata of the CD, and not in the actual contents. To protect against that, I built a new bootable CD for SD26 using the contents of the official SD25 CD, but replaced the SD25 firmware image with that from the leaked SD26 CD. Obviously, I cannot recommend using my bootable CD because you cannot verify that I am not malicious. However, I would be willing to type up the procedure for building an SD26 bootable CD if anyone is interested. This protects you against any malicious payload hidden in the metadata of the leaked SD26 CD.

The leaked bootable CD for SD26 is relatively easy to find by searching on Google for the specific file name. You will need to use the * wildcard operator because there is an additional string in the filename that is not consistent with the naming convention for the SD25 bootable CD. I got the leaked Windows executable for SD26 from this thread, before the link was taken down.

The SHA-1 checksums of the leaked Windows executable and bootable CD are:

.exe: 6ddbc4b664ab85ace4e414ccc4d7d8c2d7947f54

.iso: b7b0c7e1b9529925b0364b2cf19a62d608b58082

Update: Email sent.

[AskTheLeaf]

To make sure that the firmware update is appropriate for your drive and won't brick it, I'd recommend contacting Seagate's technical support.

It's free of charge and easy process:

http://www.seagate.com/www/en-us/about/contact_us/