HDD vertically / horizontally installed

[puntoMX]

Ow boy, a feel little kids are jumping on me here. Did I do something wrong?

[hmaster10]

My little question have cause a big ruckus among MSFN members

Thanks for all the opinions (& debates ) I was just curious coz I saw PC cases with vertical mounting position for HDD.

[eyeball]

i have an Amiga 500 with a 20mb horizontal external hard disk and it still works perfect, it must be over 20 years old now, so i dont believe gravity has an effect at all on hard disks

[meowing]

In my 29+ years of using harddisk storage media and recovering data from 'broken' disks, it is a clear-cut case for me that harddisks last longer when positioned horizontally. Of course manufacturers will state this isn't the case, but that's only because when they say it doesn't matter the HDs will wear out sooner so they simply sell more disks.

I've worked at a broadcasting facility where all the seagates that were positioned vertically were full of bad blocks and ended their lives in about 4 years of non-stop 24/7 running in servers, while the exact same disks in the exact same types of servers which had the disks mounted the old way (flat, horizontally) lasted way longer and had significantly less bad blocks on them. I built, maintained and administrated those servers, so I know this is true and way too obvious to be mistaken. Ever since the year 2000 (when those vertical disks started dying on me), I stopped believing it doesn't make a difference.

This is just my true life experience that caused me to never mount them in a vertical position again.

I don't care what anyone says, it's quite logical too if you know what the insides of a HDD look like. You would never put a vinyl record player on its side. The wear and tear would be the wrong way.

If you want your disks to live the longest, mount them horizontally, with the text-label side upwards.

Edited March 28, 2009 by meowing

[puntoMX]

Explain that to the others please, I tried it but they simple didn't get what I tried to explain.

[CoffeeFiend]

Explain that to the others please, I tried it but they simple didn't get what I tried to explain.

I dunno. I wish it was that simple really.

Is meowing's sample size large enough to make such claims? Are the "testing" conditions similar enough for this too? (exact same *everything*?) I very much doubt it, and tend to think it's more along the lines of "anecdotal evidence" myself.

it's quite logical too if you know what the insides of a HD look like. You would never put a vinyl record player on its side. The wear would be the wrong way.

Except hard drives and record players aren't quite the same! As for the record, it's not held the same way as the platters at all. As for the heads, same thing. The record player's head sits in a groove, causing wear, whereas HD heads are floating above the surface, so it doesn't cause wear "the wrong way", much like a laser doesn't wear a CD mechanically, regardless of orientation. Apples and oranges, and it explains exactly nothing (not logical).

Anyways.

Drive temps? Too hard to tell for sure. You can't easily say the temps will be better one way or another due to convection alone. Besides, airflow and case design probably make more difference here. Either ways, I'm not totally convinced temp is a such a great metric, judging by Google's Failure Trends in a Large Disk Drive Population paper (which says "Contrary to previously reported results, we found very little correlation between failure rates and either elevated temperature or activity levels.") -- not that I think high temps are really a good thing.

Mechanically speaking, it likely depends on each particular drive's mechanical design (bearings used, etc). But they're supposed to be made to work that way too (e.g. the arms are designed for it) Very hard to make broad generalizations here too.

And even then, even if it has any impact, just how much is it? The drives will have a 5% higher chance to fail after 5 years? Very hard to come up with any kind of valid or meaningful metrics here.

Also, I think there are far more significant issues. From problems with certain models (either hardware or firmware), bad designs (5 platter deathstars anyone?), bad batches, quality control problems, improper shipping/handling/packaging, etc. And of course, how it's "used" (poor case with no airflow to speak of? on a PSU with very high ripple and voltages barely within specs? vibration? PC thrown around a lot? ...) So many factors!

It's one of those you can't really prove nor disprove.

[puntoMX]

... and tend to think it's more along the lines of "anecdotal evidence" myself.

Correct, it's just what we see in the field, it's not that we have a test lab to measure temperatures for example.

[meowing]

Is meowing's sample size large enough to make such claims?

Yes. Ever since my decision to not put them vertical anymore, I've had a clear decrease in HDD problems, both in my personal life and at work.

Besides, *all* harddisks are still built on the exact same principles as the first ones that failed on me. There's not much difference between the oldest hard disk drive's structure and the newest. What was most stressing for them in the old days, is still the most stressing for them today.

Gravity will pull the pick-up-needle to one side of the drive when you put it vertical. This will *always* wear out a drive's surface as well as put extra pressure on the engine that drives the needle (that makes it move over the surface of the platters). If you don't understand that, like I wrote, look at an old vinyl record player. HDDs are built very similar to that.

http://www.youtube.com/profile?user=SuperF...deos&sort=v

Edited March 28, 2009 by meowing

[CoffeeFiend]

Yes.

Sorry, but I ain't buying it. My sample size for my "findings" is around 2500, and I still think it should be considered as anecdotal evidence. I very much doubt you have as much drives say say, Google would have, and are in a position to do a in-depth analysis of the root causes, etc. Pardon my skepticism.

Also, despite what you say, their design has evolved a great deal in the last 10 years or so (bearings, arms, heads, ...) so what was stressing/bad for them and what is now? That's most likely changed too.

This will *always* wear out a drive's surface

How would it wear the surface? Sounds like you don't even understand how it works. The heads don't actually touch the platters, and even if they did (not the case), I fail to see how it would wear it out faster than being horizontal. This claim makes no sense.

as well as put extra pressure on the engine that drives the needle

It's not an engine or a motor, it's a voice coil (an electro magnet pretty much), and that doesn't change a thing/doesn't put any extra stress on it in any meaningful way.

And no, HDs are very much unlike record players, like I said in my last post. You saying that pretty much discredits anything you said IMO.

[Maleko]

I'm going to mount my hard drives at a 45' angle now...

[Angelico_Payne]

As far as I know there is no real difference in either performance or durability.

I just know it is advisable to carry the disk verticaly or to the side of the disc horizontaly,to avoid the needle from making a sudden contact with the surface, while transporting. Dont know if it's true...

[go6o_kara]

As far as I know there is no real difference in either performance or durability.

I just know it is advisable to carry the disk verticaly or to the side of the disc horizontaly,to avoid the needle from making a sudden contact with the surface, while transporting. Dont know if it's true...

The heads are parked aside from the plates when the drive is off. So in normal case of transportation there is no risk. If you make the drive feel what's being a banana in a banana shake, well, there is no force on earth to help.