Seagate Firmware Update Bricks 500GB Barracudas

Voidsinger writes "The latest firmware updates to correct Seagate woes have created a new debacle. It seems from Seagate forums that there has yet to be a successful update of the 3500320AS models from SD15 to the new SD1A firmware. Add to that the updater updates the firmware of all drives of the same type at once, and you get a meltdown of RAID arrays, and people's backups if they were on the same type of drive. Drives are still flashable though, and Seagate has pulled the update for validation. While it would have been nice of them to validate the firmware beforehand, there is still a little hope that not everyone will lose all of their data."

Comments

At least no censoring

[amclay]

I'm glad to see them trying though. It's nice of a company to realize they made a mistake, and work to fix it.

Re:At least no censoring

[berend botje]

So far, there is no indication that they even have a QA process...

I have a solution for long term data storage.

[Anonymous Coward]

clay tablets.

If You Can Reflash It, It's Not Bricked

[Fieryphoenix]

Ay Caramba already.

Re:If You Can Reflash It, It's Not Bricked

[ResidntGeek]

The whole point of calling something a "brick" is that's how useful it is - it can't be made to do anything better, ever again. If you can plug a cable into something, and run a program on your computer that makes it able to store data or play MP3s or whatever, it's CLEARLY more useful than a brick.

Re:If You Can Reflash It, It's Not Bricked

[smellotron]

...I'm not sure "ever again" needs to be part of the definition.

Every time I've ever heard the term "Bricked", the "ever again" has been the most significant implication. The term loses its meaning if you expand it to include any device that is currently not functioning.

Re:If You Can Reflash It, It's Not Bricked

[adolf]

The definition of "bricked" depends on the ability of the speaker.

I once bricked a Linksys WRT54G. I say this because I was sure that there was nothing that I, given my knowledge at the time, could ever do to rescue it.

As time went on, I learned more about the problem. Eventually, I soldered a header to the 54G's board and built a JTAG cable, and was able to reflash its firmware more or less directly using my Gentoo desktop's parallel port. Afterward it clearly wasn't a brick anymore, since it was now routing packets just fine. I believe that the precise point at which the device stopped being a brick was between the moment when I finally understood how to repair it, and the final completion of the repair.

So, here's what I think: Given average knowledge and ability, there's lots of things that one might be able to brick. However, with sufficient knowledge and ability, nothing can be bricked.

Re:If You Can Reflash It, It's Not Bricked

[Jugalator]

Why shouldn't it be proper to say something is a brick if it can't do anything better unless it's fixed?

Then it's in need of service. You can call it "broken".

Re:If You Can Reflash It, It's Not Bricked

[Gary W. Longsine]

Uhm, no, because the term "bricked" was invented to mean broken beyond repair, not good for anything but a paperweight, like, uhm, a brick. Why on earth did you keep talking after you got that far?

Re:If You Can Reflash It, It's Not Bricked

[LateArthurDent]

The definition of "bricked" depends on the ability of the speaker.

Not really. It's bricked if it cannot be repaired by non-physical means. If you have to open the device up and start soldering leads, the device is bricked. You're just capable of unbricking it.

The term "unbrick" has been around even in the old days when "brick" was being used correctly. I think that may have been what caused the new definition to come about. People would go into forums and see things like, "I've bricked my router, anyone know if it's possible to unbrick it?" The people asking the question were looking for hardware solutions such as the one you've accomplished, but the ones new to the terminology started inferring the meaning of the term "brick" as "currently not functioning" since it was obviously possible to bring them back to life in some cases.

Re:If Seagate keeps this up

[afidel]

I wonder if this is coming from the Seagate side of the house or the Maxtor side? This sure seems a LOT more like something the old Maxtor would have done than the enterprise provider of choice Seagate.

Re:Upgrading and flashing 'untested' technology?

[fuzzyfuzzyfungus]

Arguably, when version "latest and greatest" -1 has a cool bug that causes it to permanently and (without hardware intervention) irrecoverably brick itself for no obvious reason, applying version "latest and greatest", at the manufacturer's recommendation, is a fairly reasonable thing to do.

Anybody who thinks that RAID=backup is going to learn an exciting lesson; but I don't think we can, in fairness, blame people for applying the update.

Oh what a long, long fall.

[jd]

Once upon a great while back, Seagate was one of the première names in hard disk technology. These days, the only press I'm seeing them get is bad firmware, questionable reliability, etc. They've been around longer than Microsoft, they really have no excuse at this point for not even testing their bugfixes on their own hardware. It's not like they even have to test third-party stuff.

What leads to this sort of decline?

Re:Oh what a long, long fall.

[tono]

I don't know, you should ask IBM the same thing. Deathstar?

bad Seagate, bad!

[Eil]

I work for a web hosting company and we get these drives by the case. I couldn't guess how many are deployed throughout the datacenter but on some of our backup servers alone I've calculated that I have almost 100 drives that need the firmware update. Thankfully none of the disks on the systems that I admin have shown problems yet, but we try to run a quality operation and that includes preventive maintenance wherever possible.

I was all set to update the firmware on these when one of our guys found that the update rendered unusable 8 of the 8 drives he upgraded the day before Seagate pulled the update. We currently have some massive amount of Western Digital 500GB and 750GB disks on rush order as a result of this debacle. It wouldn't surprise me if management tells us to swap the Seagate disks for the WDs and decides to just sell the whole lot of Seagate disks off in bulk as defective. It would be cheaper than paying people to update each one by hand.

Before this, Seagate used to mean "quality" in my opinion as their failure rate seemed to be lower than the competition and their 5-year warranty was unmatched. For the average home user, this situation is a headache. For people running datacenters filled with these disks, it's an outright fiasco.

Not bricked!

[ZorkZero]

It's not bricked if you can fix it without modifying the hardware. It's a nice term -- stop destroying it.

Re:THE FACTS

[rossz]

There's a lesson to be learned here. DON'T FARKING LET MIDDLE MANAGEMENT BYPASS YOUR TRIED AND TRUE TEST/RELEASE PROCEDURE. Yes, the initial problem was bad, but the rush to get a fix out made it much much worse. Upper management is at fault here for allowing middle management pencil pushing idiots to do this to the company's reputation. Procedures are in place for a damn good reason.

Re:THE FACTS

[maxtorman]

You say that now, but you have to admit, with such screaming and carpet-clawing that went on about the 1.5Tb issue, some of the fault rests on the mob mentality pushing Seagate management to get a fix out ASAP for an issue recently proven. I'm not saying it's okay - but the exact same situation that can force a large and lumbering company to move faster, can force management to push really hard and cause quality systems to break down. You can whip the bull to get it to run, but you may just cause it to run right off the cliff. :)

Bricked Threshold

[raehl]

A brick's value is the cost of creating a brick to replace it.

So if it is less expensive to throw something out and buy a new one than it is to repair it, it's bricked.

Re:Pwnt.

[pipatron]

Raid has never been a backup. A backup is something stored outside of the running set. That way you can restore the data if your running system would, you know, break down.

firmware update for drive = fail

[stiller]

Ok, maybe it's just me, but who the hell updates drive firmware anyway? Just because I'm a techie, doesn't mean I am suddenly willing to do more work than other customers.
Do you think a single consumer out there goes through the trouble of updating their drive firmware? (unless there's an automatic procedure in place, like probably mac and some windows manufacturers have)

To me, any drive which requires an firmware update to function (not just perform better) after purchase, is a failed product and I would surely hesitate to buy another ever again.

I used to buy Seagate drives in pretty large numbers for some of my datacenter activities and every time a drive locked up for some reason, I insisted on a new drive through EMA. Had Seagate refused, they would have taken away a large chunk of their added value, to me. I would probably never buy another drive from them again.

Re:Pwnt.

[Kokuyo]

It is the old High Availability versus Disaster Recovery question. Two completely different things aimed at two completely different problems.

The first is to make sure that your system remains available as long as possible even if some of your hardware goes belly-up. The latter is for when your DATA goes belly-up.

Why not include base firmware on a rom?

[jollyreaper]

I never understood why equipment capable of being flash-updated by users does not include the 1.0 drivers as a ROM onboard the device. This way if you completely and utterly bork the flashing, you can reset a jumper, press a recessed button with a paperclip, so SOMETHING that will cause the EPROM to be reflashed from the known good ROM. "Hey, here's baseline firmware again, people. Let's try this again."

The only possible explanation I can think of for not doing this is that the known-good ROM would add another half-cent to the manufacturing process and we know how manufacturers watch their pennies.