Reliability of HDDS in adverse environments

[Lycidas]

Hey guys, I'm planning on doing a project where I record large amounts of data in an environment that would vibrate. The oscillations would be pretty small most of the time with an occasional burst of a few violent vibrations. I was curious as to the reliability of a physical disc drive in this kind of an environment.

I'll be recording huge chunks of data and then erasing them as they become undesired after a period of time (which explains my desire for a HDD over a SSD). Cost is also somewhat a factor here too which is why I wanted the HDD.

So does anyone have any knowledge of this like running a computer that's attached to a car or some kind of moving vehicle perhaps? That's probably the closest environment to what I'll be doing that I can think of.

Thanks!

 

[drescherjm]

How about raiding a few laptop hard drives? And regardless of what drives you choose have a backup strategy.

 

[OldSchool]

My room mate has an industrial client who has a workstation on their factory floor next to a huge tumbler machine, it is part of the control/monitoring apparatus. He said when they were using mechanical drives in it that they would be impressed if one lasted longer than a few months (they had cloned drives on site to switch out). I believe they have switched over to a RAID 1 of SSDs and there have not been any issues so far.

 

[Lycidas]

He said when they were using mechanical drives in it that they would be impressed if one lasted longer than a few months

This is kind of what I was expecting but not wanting to think of. I may have to settle for some kind of SSD, but there's going to be a lot of write/erasing going on which in general isn't a great thing for SSDs to deal with...

Since my vibrations won't be nearly as violent as a tumbling machine, the HDDs that I'd use would probably last a year, which still isn't a very good lifetime.

How about raiding a few laptop hard drives?

Good point, laptop drives would probably do better than regular drives when it comes to vibration since laptops are expected to move constantly. Thanks!

 

[lovemyram4x4]

I used to drive around w/ my laptop on the whole time(I did this for several years) and never had a HDD fail. I do field server work so I did a lot of drive over 50k/yr and the laptop would sit on one of those metal swinging monitor stands(some other used now swinging stands) so it was like it was sitting on a nice cushy seat.

 

[Lycidas]

I used to drive around w/ my laptop on the whole time(I did this for several years) and never had a HDD fail. I do field server work so I did a lot of drive over 50k/yr and the laptop would sit on one of those metal swinging monitor stands(some other used now swinging stands) so it was like it was sitting on a nice cushy seat.

I don't think I could filter out the vibrations in anyway, so unfortunately that won't really be an option. Even a stand would be forced to experience oscillations. I had originally thought about using spring/damper systems to see if that would help but I'm not quite sure of how effective that might be. Thanks for the input though!

 

[bernz]

Use elastic suspensions. You can rig up something that's held in place by bungee elastic or something: I did this all the time to mechanically de-couple a hard drive from a case to reduce vibration noise. No reason why it won't work the other way around. Check silentpcreview's forums for hdd suspension tips

 

[sethmo]

Use SSDs or a think client. You can get XP embedded thin clients without HDD if you dont have a terminal server.

 

[lovemyram4x4]

I don't think I could filter out the vibrations in anyway, so unfortunately that won't really be an option. Even a stand would be forced to experience oscillations. I had originally thought about using spring/damper systems to see if that would help but I'm not quite sure of how effective that might be. Thanks for the input though!

My point was the laptop was subjected to plenty of vibration and shock(as the stand could swing-hard braking) and the laptop drives held up.