Seagate: An Admission of Guilt

[Zorb]

Finally, they own up to one of their big issues... Even made Slashdot!

http://hardware.slashdot.org/article.pl ... 17/0115207

Seagate offers to recover the content for free in case of this issue, however, I've been observing media problems with the 7200.11, so we shall see what comes of this, if they take responsibility for that too.

I know that with the 320GB versions of both families, I have observed more than ten times the failure rate in four months of drives I have sold.

9/58 ST3320613AS failed in 6 months.

3/172 ST3320620AS failed in 6 months.

My experience from records.

[Doomer]

I saw a lot of WD drive with bad heads
should I request free DR from WD?

[Steve]

It's not IF a hard drive will fail , but When.

As BlackST would say
PERIOD

[Zorb]

Finally, they own up to one of their big issues... Even made Slashdot!

http://hardware.slashdot.org/article.pl ... 17/0115207

Seagate offers to recover the content for free in case of this issue, however, I've been observing media problems with the 7200.11, so we shall see what comes of this, if they take responsibility for that too.

I know that with the 320GB versions of both families, I have observed more than ten times the failure rate in four months of drives I have sold.

9/58 ST3320613AS failed in 6 months.

3/172 ST3320620AS failed in 6 months.

My experience from records.

I really must have been asleep when I wrote this...

The ST3320620AS is Barracuda 7200.10
The ST3320613AS is Barracuda 7200.11 and is notably worse.

As for the if vs when for failure, of course a drive will fail. It's a mechanical device with moving parts. We all know that. Either way, since I have been having these issues with 7200.11s, I switched to Hitachi and have had far fewer problems in all my internal metrics. Long live the deathstar.

[Doomer]

Either way, since I have been having these issues with 7200.11s, I switched to Hitachi and have had far fewer problems in all my internal metrics. Long live the deathstar.

It would be interesting for you to know that the term "deathstar" came from disks produced by IBM/Hitachi.
It started with DTLA family and ended with AVVN family which had horrible quality
and it named "deathstar" because of "DeskStar" products which actually still can be found on Hitachi site

So, yes, long live to your Hitachi deathstars

[Zorb]

The AVVA and AVVN didn't have nearly the problems that the DTLA and AVER did.

Either way, they have improved to be what they were before those few models, and as such are one of the more reliable drives in my experience, despite quality problems with the models in question.

[Doomer]

The AVVA and AVVN didn't have nearly the problems that the DTLA and AVER did.

Yes they did

[guru]

Most of the problems were related to the Hungary factory from what I remember. Wasn't it the glass media that was causing the problems? It's very hard to get perfectly flat glass media.

Anyway most drives are sh1t and thats why we are in business!!

[scratchy]

It's very hard to get perfectly flat glass media.

Its also very hard to get anything to stick to it properly too. When the drives heat up, the glass expands ata different rate to all the other coatings resulting with microcracking, blisters and eventually nice concentric stripes in the media.

<itch>

[Zorb]

Right, if they had good cooling, especially shallow temperature gradients, they didn't have this sort of problem nearly as much. The other problem was that the material would break away (think of flaking paint) from the point of damage, which is one reason almost transparent Deskstar platters have been seen. This was worsened by the fact that some firmware versions of the DTLA and AVER drives were programmed to keep the heads near the center of their sweep, for faster average access to information (it only really helped with synthetic benchmarks). This was changed, to implement an automatic sweeping between extremes, in the later versions, and with a firmware update to older drives.

Regardless of what Doomer said (I do respect his information and he may be right on the whole), my observation showed a decrease in failure rate in later models, (AVVA/AVVN, AVV2, in order), though the AVER was worse than the DTLA.

Like I said, I used to ship the 120GXP(AVVA) and 180GXP(AVV2) on my mid level workstation class machines (when I still sold a lot of hardware), and I used to mount them in 5.25" aluminum removable racks, which had two 30mm fans in the front of it and one in the back, and I never had many problems with drives. When I started seeing an unusual number of other machines in for repair with failing or dead drives of the other families, I switched my own line to Seagates (Barracuda ATA V at the time). I never sold the DTLA and AVER drives, I sold Barracuda ATA III, IV before switching to IBM. I actually still have some of those old machines that I still service (back from the time I was a one person company with part time help from friends), and most have their original drives, but I'm sure most of the others are retired (P4 1500-2000 or P3 1133 only gets you so far).

[Doomer]

Wasn't it the glass media that was causing the problems? It's very hard to get perfectly flat glass media.

Apparently not
Many modern 2.5" drives use glass media
I personally saw Fujitsu MHT with glass media broken in pieces

There were two major problems with IBMs: PCB contacts problem and R/W channel glitch problem
Both problems caused improper heads movement which caused media damages

PS: and Hungary factory has nothing to do with it. It was just speculations

PPS: Hitachi still doing drives which migth have MD

[guru]

Well most of the failed drives I had were from Hungary. Maybe that was down to geographic locations within the distribution channel.

Sure it's easier to get flatter media on a 2.5" diameter HDD.

I have some reports from IBM on the matter. probably can't post it here....

[Zorb]

MD?

[Doomer]

MD stands for media damage

[Zorb]

That's what I thought.