Seagate 3TB lawsuit

[Heihachi_73]

Re: Seagate 3TB lawsuit

Meanwhile my 40GB ST340015A is still going strong after 80000 hours of 24/7 use (82736 hours according to SpeedFan's SMART readout). My trash-picked 60GB ST360015A isn't much further behind at 72460 hours.

[momaka]

Re: Seagate 3TB lawsuit

^ Yeah, the old Seagate HDDs (Barracuda ATA-IV and 7200.7, IIRC) are extremely reliable. Now, the interesting bit is that many of them would actually develop a few bad sectors here and there as the drives got filled up over time. But once they stabilized (i.e. all of the bad sectors were remapped or left pending), the drive no longer would have any issues after that. At least that was the case in my experience. I have surveyed only about 30-40 of these drives, though.

[ChaosLegionnaire]

Re: Seagate 3TB lawsuit

hmm thats interesting. first time im hearing its normal for hard drives to have bad sectors. in my experience, once a hard drive starts having bad sectors, it starts to "spread" and keeps increasing. drive is trashed and can no longer be used to store critical data or an os. only maybe as a tertiary backup drive.

[momaka]

Re: Seagate 3TB lawsuit

hmm thats interesting. first time im hearing its normal for hard drives to have bad sectors.

Well, it shouldn't be .
If you think about it, we live in an imperfect world. Therefore, no matter what we do (in terms of manufacturing), those HDD platters will not always have a perfectly uniform magnetic coating, meaning some areas may be weaker than others. When the HDD is new, perhaps the magnetic coating is good enough, and it all works well. But as the HDD ages, some of the weaker areas on the platter could go bad enough to cause data corruption. And these areas could be quite big (thus causing many bad sectors) or just tiny areas affecting only a few sectors. Moreover, these areas may become even weaker over time, causing more bad sectors. Or they may become weak, cause some bad sectors, then settle down and not degrade significantly to cause any more bad sectors.

Based on all of that, there are four (4) possible distinct scenarios:

a) the HDD develops a few bad sectors, then settles down and no longer develops bad sectors (or so rarely, that the HDD will become functionally obsolete before it breaks down)

b) the HDD develops many bad sectors, then settles down and no longer develops bad sectors (or so rarely, that the HDD will become functionally obsolete before it breaks down)

c) the HDD develops a few bad sectors but never settles down, thus continuing to develop bad sectors over time

d) the HDD develops many bad sectors but never settles down, thus continuing to develop bad sectors over time.

From these scenarios, a) and b) could *potentially* lead to a good stable HDD once it is done mapping out all of the bad sectors (two passes of zero-fill / low-level format will tell you everything you need to know ). Scenarios c) and d) is what you should avoid (especially d), because those HDDs may never stop developing bad sectors and could go out at any moment once the ROM on the HDD's platters gets affected.

Also, just because an HDD has 0 (zero) bad sectors doesn't mean that it's good. The HDD could still jump straight into scenario d) above and fail within minutes/hours/days/months. However, if the HDD has had 0 bad sectors over a very long time in use (and with a lot of data written), then chances are it has good coating on the platters and it is unlikely to start developing many bad sectors.

That said, all of this above only considers just the magnetic material on the platters. You could have an HDD develop bad sectors due to head going bad or out of alignment, actuator arm or spindle motor bearings getting sloppy and loose. With those, however, the HDD will most likely go straight into scenarios c) or d) .

Finally, it should be noted that those scenarios above are not clear-cut. An HDD may fall between any of them. So if you really need to be 100% sure your data won't go corrupt, it is best to avoid using an HDD that has non-zero bad sectors. Better yet, pick one that has had a few years of use (but not too many) AND zero bad sectors.

But the best solution: always make back-ups (multiple!)

[fzabkar]

Re: Seagate 3TB lawsuit

I've been watching professional data recovery forums such as HDD Guru for over 6 years. My impression is that bad sectors in modern drives are very (most?) likely due to head faults rather than actual media problems. AIUI, the low fly height (4nm or less) means that modern drives are much more likely to experience head slaps and thermal asperities. These cause permanent degradation of the head characteristics, resulting in "weak" heads.

The risk with weak heads is that the large number of "bad sectors" quickly overflows the storage space assigned to them in the G-List (grown defect list). When this happens the drive goes offline and no longer appears in BIOS. Furthermore, those other firmware modules that need to be updated on a regular basis (eg SMART, translator) become damaged. This is because these modules are stored in a reserved area of the platters (System Area, Service Area, Maintenance Cylinders, Negative Cylinders, etc), not in flash memory on the PCB. If the head becomes weak, then writing (and reading) these firmware modules becomes unreliable. The result is that the drive cannot load its own firmware during its own POST.