HDD Power Settings Concerns

[Collider]

So I have 4 mechanical and 3 SSD drives in my desktop. OS runs on one of the SSDs, rest of the mechanical drives are used for games, storage, backup, etc.

With Window 8's default power settings (20 min) I hear my 4 mechanical drives go off-line one by one producing high pitch noise.

I'm OK w/ the noise and the occasional frustration of accessing a file and having to wait an extra second or 2 while the drive spins up.

My concern is the longevity of drives. By having them frequently spin up/down wouldn't that detract from their lifetime vs. just having them on?

 

[Elixer]

Yeah, it would.
Might be better to purchase some hotswap bays, and only turn on the HDs when needed. It also makes swapping out HDs really easy for backups.

 

[thecoolnessrune]

As Elixer said, that on/off switching is one of the primary reasons that hard drives fail in consumer devices. If you need access to the files on a regular basis, then disable power management on the drives. Modern hard drives do not use that much power and as long as they are kept ventilated will last longer this way

 

[KingFatty]

It depends. You may have the perception that the drives are switching on/off all the time. But it may be that switching them off saves them from failure because they are spending less time turned on.

Also, the drives have been designed to be able to turn on and off without stressing themselves out. Consider drives like the Green series, where it's hard-coded into the drive itself to spin itself down. It's good for the drive to avoid generating heat and wear, and in Green drives, it happens at a basic level where even the OS can't stop it from happening. I think it's good that the OS lets you emulate that behavior in drives like yours (assuming they are non-green drives). Why do you think Windows has this default setting, a way to break your system faster so you have to upgrade? Or because it's generally recognized as a good thing for drives?

Also, in studies that try to analyze hard drive failures, it seems that heat and hours of operation are the big factors in determining whether a drive fails. The number of spin-ups and spin-downs did not seem to be a major factor of failure. Spinning-down the drive conserves heat and hours of operation, so it's doubly beneficial for longevity.

 

[Elixer]

Also, in studies that try to analyze hard drive failures, it seems that heat and hours of operation are the big factors in determining whether a drive fails. The number of spin-ups and spin-downs did not seem to be a major factor of failure. Spinning-down the drive conserves heat and hours of operation, so it's doubly beneficial for longevity.

Which study do you speak of ?

I know quite a few people that had the green drives that kept spinning down/up very frequently die in under one year.
Temps didn't play a role at all.

 

[Collider]

thanks for the replies - set power options to 'hdd always-on'.

any recommendations on PC sleep setting, also component longevity-wise, as is there any additional stress to components (don't care much for power savings) ?

 

[KingFatty]

Which study do you speak of ?

I know quite a few people that had the green drives that kept spinning down/up very frequently die in under one year.
Temps didn't play a role at all.

http://static.googleusercontent.com/external_content/untrusted_dlcp/research.google.com/en/us/archive/disk_failures.pdf

Would you say the green drives are designed to fail?

 

[C1]

At one time I counted 34 drives on hand (always said that I had a bushel of em). Ive since gotten rid of a bunch, but still have many. I have a now 10 year old P4 system I use every day with four drives (2 x 120 WDs and 2 x 300 Maxtors) plus I always have two USB notebook drives connected. All these drives hooked to the big machine eventually spin down when not in use and can be heard to spin up when accessed. Have a bunch of Hitachi 2TBs too. They also spin down after 9 minutes of non-access. Also have a couple of notebook computers I use daily. I also used lots of different computers in the many years while working.

I dont recall having lost more than two drives in my whole life (both were 3.5").

What the hell is going on with you people and drive reliability?

A guess is some kind of abuse or buying shitty support hardware (eg, crappy power supplies).

PS: Over the years now I have taken to watch movies via my computers. Hundreds of movies so far and some of those drives must now have thousands of hours on them.

 

[Elixer]

http://static.googleusercontent.com/external_content/untrusted_dlcp/research.google.com/en/us/archive/disk_failures.pdf

Would you say the green drives are designed to fail?

Ahh yes, the google study, which is really out of date now. They need to do this yearly.

I think all drives made today are inferior to what they were.

At one time I counted 34 drives on hand (always said that I had a bushel of em). Ive since gotten rid of a bunch, but still have many. I have a now 10 year old P4 system I use every day with four drives (2 x 120 WDs and 2 x 300 Maxtors) plus I always have two USB notebook drives connected. All these drives hooked to the big machine eventually spin down when not in use and can be heard to spin up when accessed. Have a bunch of Hitachi 2TBs too. They also spin down after 9 minutes of non-access. Also have a couple of notebook computers I use daily. I also used lots of different computers in the many years while working.

I dont recall having lost more than two drives in my whole life (both were 3.5").

What the hell is going on with you people and drive reliability?

A guess is some kind of abuse or buying shitty support hardware (eg, crappy power supplies).

PS: Over the years now I have taken to watch movies via my computers. Hundreds of movies so far and some of those drives must now have thousands of hours on them.

Ever since the floods, HD quality has been iffy at best. You will find tons of drives that are DOA or fail before year one on pretty much any site that sells HDs.
Does this have to do with crappy packaging, or does it have to do with cheaper quality parts being used ?
Nobody can say for sure.

If you have been around these forums much, you would see that people who repair HDs for a living are cursing the "green" drives (both WD & seagate), and are not fond of seagate either.

Even WD with the red label drives, supposed to be used 24/7 are showing up with lots of dud units.

True, we don't have hard numbers to see what actually is going on, but overall, the quality post flood days, HDs have taken a nose dive.

 

[KingFatty]

Ironically I just yesterday had a drive fail on me. I wanted to find out what happened. The symptom was the drive was making a tone on/off/on/off, though the tone was generated by a buzz sound like the motor trying to spin.

When I disassembled it, I saw the arm was frozen half way along the radius. The head was stuck on the platter. So I think it was a head crash, and the motor couldn't spin the disk or move the arm so it made that buzzing sound that an electric motor makes when it tries to go but is held still.

This kind of crash, I think would be from just bad luck, but I wonder if it was related to spinning up or spinning down? It's a laptop hard drive that I would make a backup on, and put in a fireproof safe shared by the wife. Parts of the drive were a bit banged up so I think it was jostled around in the safe, so maybe it was that.

Anyway, I was also surprised to see the disk platter could be shattered like a CD, except it seemed more like glass than plastic. Previously, the last hard drive I disassembled, was much older and it had metal platters that I could bend.

Anyway, I just wish there was a study that could build on the google study, and actually disassemble the failed HDDs to check for *how* they failed. Was it just electronic failure, or motor failure, or head crash, or what?

But I think if my drive was destined to fail with a head crash, it was just a matter of putting enough hours on it, because I don't know if head crashes are brought about by spin-up/spin-down of the platter. Instead, I think it's just normal back/forth movement of the arm. So maybe here it wouldn't have made a difference whether the drive was left on spinning, because I think in that situation the arm doesn't move while the platter spins? Although, if I did leave the drive on, I think windows would occasionally move the arm now and then, so technically maybe it would have added more likelihood of the head crash compared to powering off the drive where I know the arm won't move?

 

[C1]

All right then. I can appreciate what is being said. Maybe HDDs simply need to be dealt with now as throw away.

Here is an interesting article that sheds light on drive reliability that is not too old :
http://blog.synology.com/blog/?p=1130

 

[thecoolnessrune]

What the hell is going on with you people and drive reliability?

A guess is some kind of abuse or buying shitty support hardware (eg, crappy power supplies).

Please tell that to NetApp, EMC, IBM, Hitachi, and all the major storage vendors where our number one replacement is hard drives due to failure. In our Datacenters, we replace at least 1 failed hard drive a day. This is why storage vendors have extensive support contract systems to rapidly replace these failing drive. Your opinion that hard drives rarely fail is that of delusion and anecdotes.

 

[Concillian]

Technically yes it will reduce the life of a drive, but so will running 24/7...

Realistically this failure type stopped being a primary failure type somewhere between 10 and 15 years ago. How do I know? I have worked in various areas of hard drive manufacturing and R&D for 16 years and know how designs have changed over the years.
I don't really work in an area that has any kind of complete drive failure reports, I'm pretty far upstream of that kind of stuff, but the designs have changed dramatically from the days when you could sometimes fix issues by tapping a drive with a hammer when it was trying to spin up to get the heads unstuck from the media (15-20 years ago.) The current designs are such that the heads don't even touch the media anymore when the motor spins down, making the only issue with spin up / down mechanical wear of the motor and such... which also occurs with the drive spinning.

Go ahead and spin them up & down.
15 years ago I kept my drives running 24x7. Today, I spin them up and down.

 

[Red Squirrel]

Yeah definitly leave them running. I'd say most of the drive failures I've seen myself have happened after a cold boot. I have about 200AH worth of battery backup for my servers to prevent exactly that. The few times where I had to shut stuff down I usually lost a couple drives when I turned it back up after the power restored.

If anything drives these days should be in a raid array and not stand alone given their sizes and low reliability, and you definitly don't want to spin down drives in a raid.

 

[Dessert Tears]

Might be better to purchase some hotswap bays, and only turn on the HDs when needed. It also makes swapping out HDs really easy for backups.

I've been thinking of getting a bay for both of these purposes. Any recommendations? The SNT SNT-125B has great reviews and is cheap, but it went out of stock at Newegg and may have been discontinued.

 

[Red Squirrel]

I have two of these in my server, they're pretty nice:

http://www.newegg.ca/Product/Product...82E16817121405

5 sata ports on the back. They are not trayless though.

This one is trayless:

http://www.newegg.ca/Product/Product...82E16817990012

I owned the 4 bay version of those too.

 

[Dessert Tears]

The SNT SNT-125B has great reviews and is cheap, but it went out of stock at Newegg and may have been discontinued.

It was discontinued, but I grabbed one on eBay. I added a cardboard shim to have the drive sit in the tray without screws. I occasionally use a cracked-open enclosure that's about as sketchy.

The iStarUSA T-7M1-SA is much nicer and has screw holes for 2.5" drives, but it starts around $35 depending on color.