How can I maintain HDD to increase its lifespan?

[rOadToIS]

How can I maintain HDD to increase its lifespan and to prevent unexpected failure?

[AKAJohnDoe]

Originally Posted by rOadToIS

How can I maintain HDD to increase its lifespan and to prevent unexpected failure?

Expect failure and backup frequently

[jonyjoe81]

Put a fan on it. On my mediapc which runs 24/7 with 4 hard drives on it, each hard drive has it's on fan blowing on it.

I've had that computer running 24/7 for 3 years recording TV all the time, I've only lost one hard drive a 40gb boot drive (that was already 3 years old and had no fan). Since I've been putting fans on all my hard drives, I haven't lost any drives from failure..

I'm not sure if the fans have anything to do with my hard drives lasting a longtime but it's something that seems to work for me. Everytime I build a computer, I always make sure I have enough space between hard drives so the fans can do there work. Don't put one drive on top of the other.

I think 45 celcius is the "sweetspot" for a hard drive, but anything lower than that is even better. Without a fan my hard drives can reach 50-55 celcius real fast.

Even on my externals I put a room fan blowing on them when I use them for an extended period.

[YeOldeStonecat]

Purchase higher quality models with longer warranties and higher MTBF ratings.

All the brands out there sell generally 3x classes of hard drives..
*Budget lower quality models with 1 year warranty, and usually MTBF below 500,000
*Mid-range models, 3 year warranty, MTBF above 600,000
*Enterprise Class...5 year warranty, MTBF of 1,000,000 to 1,200,000

And yes..keep it cool. Most chassis these days will have air intake vents in front of the chassis, so that the exhaust fans of the chassis pull air in through that...thus across the hard drives...on the path towards the exhaust.

I have an older Antec 1020 server case with a large hard drive bay..that has fans in front of the drive cage. My current gaming rig is an Antec Sonata.one large exhaust fan, the air intake vent is right in front of the hard drive cage..so it always has cool air incoming that goes right across the drive.

As stated above, I also try to avoid "stacking" drives right next to each other...if you need multiple drives in a computer case..try to skip a mount between each drive, so you have that roughly 1" of space between drives.

And in the years of working in IT, I've noticed a relationship between failed drives...and when I open the case..the case and often the drive is covered with a layer of dust bunny blankets. Preventing air from properly cooling the drive. I have also noticed that drives mounted directly on top of each other will have a higher rate of failure..they can't keep cool as easily.

[TechOutsider]

My Maxtor Fireball was amazingly reliable; lasted over 5 years in a poorly ventaliated chassis with a P4. And it scored a 5.3 on Vista. And it's only 40gb and 5400rpm.

My other hard drive in my 10 year old Dell XPS still works, on 98, 2k, or even XP.

Is there a way to find how many hours your drive has been used?

[GlobalForce]

Gee, thanks for taking the time to post people, your replies have really been helpful.

As a matter of fact, beyond a sole instance you haven't bothered to thank anyone, anywhere, your entire first page of thread starts. What gives, you above a response? They don't practice manners where you're from, reciprocation a somehow obscure concept? ~shakes head .... Terrible!

S

[Dark Shadow]

Originally Posted by GlobalForce

Gee, thanks for taking the time to post people, your replies have really been helpful.

As a matter of fact, beyond a sole instance you haven't bothered to thank anyone, anywhere, your entire first page of thread starts. What gives, you above a response? They don't practice manners where you're from, reciprocation a somehow obscure concept? ~shakes head .... Terrible!

S

10 X Global

[pugmug]

Originally Posted by GlobalForce

Gee, thanks for taking the time to post people, your replies have really been helpful.

As a matter of fact, beyond a sole instance you haven't bothered to thank anyone, anywhere, your entire first page of thread starts. What gives, you above a response? They don't practice manners where you're from, reciprocation a somehow obscure concept? ~shakes head .... Terrible!

S

Agree 100% with your post.You do realize that dealing with children and the way parents raise them these day's, manners are a thing of the past.

[Long View]

Originally Posted by jonyjoe81

I think 45 celcius is the "sweetspot" for a hard drive, but anything lower than that is even better. Without a fan my hard drives can reach 50-55 celcius real fast.

Hi jonyjoe81 - are you living somewhere hot ? I have a number of machines and none ever get that hot. Right now I'm reading 24 c and 24 c for C: and D: with ambient temp of 63.1 F or 17.3 c Each machine has a 120 mm fan at the front and another to extract. Each drive sits in a Zalman cooler
http://tinyurl.com/by2nkv

[markymoo]

less powerups. keep your pc on for longer. turning hard drive on and off alot wears it more. not pratical but still.

[AKAJohnDoe]

Folks, it is a generic tool. A hammer or screwdriver. Replace it when it wears out. The data is what needs maintenance, ala backups.

[MrBrian]

Originally Posted by TechOutsider

Is there a way to find how many hours your drive has been used?

Look at 'Power-On Hours' S.M.A.R.T. attribute. Any program that can read S.M.A.R.T. data should be able to do this. HD Tune (free) can handle this.

[Mrkvonic]

Hello,

I think constant temperature is more important than the actual value itself, except the very low-end and high-end ones. But if your drives pur at anywhere between 25-50 degrees C or so, constantly, they're in a steady state, which is good for electronic parts.

The annual fluctuation in drive temperatures I have between seasons, from winter to summer, is no more than 5 degress C and they're never turned off except very rare cases, so I think this makes them feel 'good'

Mrk

[Kerodo]

Originally Posted by Mrkvonic

Hello,

I think constant temperature is more important than the actual value itself, except the very low-end and high-end ones. But if your drives pur at anywhere between 25-50 degrees C or so, constantly, they're in a steady state, which is good for electronic parts.

The annual fluctuation in drive temperatures I have between seasons, from winter to summer, is no more than 5 degress C and they're never turned off except very rare cases, so I think this makes them feel 'good'

Mrk

Yep, this seems to be my experience also, whenever I run my drives 24/7 and rarely shut them down, I get very good life out of them, 5-7 years usually. I think they're happier always running at a constant temp and so on also.

Good clean power supply via a UPS or a reputed surge strip as well as good quality PSU and keep the temps down. I got ancient IBM IDE and SCSI, both around 9 years old still running till today.

[FastGame]

Originally Posted by markymoo

less powerups. keep your pc on for longer. turning hard drive on and off alot wears it more. not pratical but still.

Really ?

Keep your drive cool and turn the PC off when not in use, thats the secret

MTBF rating = the longer it runs the sooner you get there.....

[noone_particular]

You'll see 2 opposing schools of thought on this subject, both of which are partly true. Running a PC 24/7 does help to increase the longevity of its components, both mechanical and electronic. With internal combustion engines, most of the wear takes place during startup. The same holds true with hard drives. Hard drives that run continously will last for more operational hours than one that's operated intermittently. Whether that translates into a longer lifespan in years depends on too many factors.

My PCs run 24/7. I've had 2 hard drives fail in the last 6 years. Both were obtained used. Both were nearly 10 years old. This PC (HP Pavilion 4463) will be 10 years old this year. Except for occasional cleaning, it's run 24/7 for the last 5 years. It's worn out 2 floppy drives and a CD burner. The original 5GB hard drive is still going strong, as is all the rest of the original hardware. I also have a Gateway P5-133 that's been converted to a hardware firewall, Smoothwall 2.0. It's run 24/7 for the last 3 years. Still has the original hard drive.

[FastGame]

Originally Posted by noone_particular

You'll see 2 opposing schools of thought on this subject, both of which are partly true. Running a PC 24/7 does help to increase the longevity of its components, both mechanical and electronic. With internal combustion engines, most of the wear takes place during startup. The same holds true with hard drives. Hard drives that run continously will last for more operational hours than one that's operated intermittently. Whether that translates into a longer lifespan in years depends on too many factors.

Your "internal combustion engines" vs electric motor analogy is rather weak....internal combustion engines wear during startup is due to the lack of lubrication, something the electric motor doesn't suffer.

The HD makers list Cycle times (startup/shutdown) for the various drives, in all likelihood the PC case switch would wear out from the on/off before the HD would.

Heat is the #1 danger to electronics, the longer its exposed to heat the greater the risk for failure.

BTW, if anyone wants to use the Server analogy we could discuss the operational differences between Servers and Desktops.

[noone_particular]

Your "internal combustion engines" vs electric motor analogy is rather weak....internal combustion engines wear during startup is due to the lack of lubrication, something the electric motor doesn't suffer.

Heat is the #1 danger to electronics, the longer its exposed to heat the greater the risk for failure.

Lubrication is a factor in electric motors as well, though not to the same degree. As long as the motor is running, lubricant is evenly distributed. Temperature and lubrication are interacting. Lubricant is more effective at operational temperatures than at room temperature.

Regarding heat, as long as the temperature remains within limits, heat isn't a major factor. It's only when it becomes excessive that actual temperature comes into play. Continous operation has an advantage here in that there's less temperature variance. Operational temperature is significantly higher than room temperature. Clearances change as temperatures vary as does the effectiveness of the lubrication. In 24/7 operation, temperatures, clearances, etc reach an equilibrium and remain nearly constant. These are the normal conditions the drives are designed to operate at. More wear takes place when conditions are not at these design parameters, like those found during startup when bearing sufaces are cooler and a bit tighter, lubricant is stiller, resulting in a bit more friction and localized heat. Temperature variation, and all the changes that happen because of it are a factor almost as much as absolute temperature, provided that the absolute temp doesn't exceed operational limits. There's also the factor of thermal shock to electronic components during power up. Electronics, including those in the hard drives, fail more often during power up than at any other time.

[FastGame]

Originally Posted by noone_particular

Clearances change as temperatures vary as does the effectiveness of the lubrication. In 24/7 operation, temperatures, clearances, etc reach an equilibrium and remain nearly constant. These are the normal conditions the drives are designed to operate at. More wear takes place when conditions are not at these design parameters, like those found during startup when bearing sufaces are cooler and a bit tighter, lubricant is stiller, resulting in a bit more friction and localized heat.

The bold print needs to be fixed, Heat= expansion and Cold= contraction, there's no way bearing/spindle fit is "a bit tighter" during startup.

As for the points you make there's not much argument from me, I don't believe in perpetual motion....I do understand the HD's operational design specifications.

Now

Why do the 24/7 people want to ignore one of the HD's (known, tested, minimal) operational design specification ? Start/Stop Cycles

Example

IBM drives that use its head load/unload technology are often given a specification for minimum load/unload cycles instead of start/stop cycles. This is basically the same concept, except that the numbers are typically much higher. Some IBM notebook drives are spec'ed for 300,000 load/unload cycles! Even if you started and stopped the drive 100 times a day it would take you over eight years to get to that number.

Does Start/Stop Cycles take into account all the things noone_particular speaks of ? yes it does.

One thing I forgot to mention in my recommendation for longer HD life, use Linux. For the 24/7 people, Linux has far less HD activity during operation than Windows.

A new can of worms

[Kerodo]

In the end, I think most of this is just talk, and all we can do is speak from actual experience. I run mine 24/7 and they have all lasted 5-7 years doing that. On the other hand, I just bought a new PC and the HD crapped after 2 months and died. So **** happens. But.... my experience tells me that running 24/7 works out well.

[MrBrian]

Google’s Disk Failure Experience

The Googlers conclude:

In the lower and middle temperature ranges, higher temperatures are not associated with higher failure rates. This is a fairly surprising result, which could indicate that data center or server designers have more freedom than previously thought when setting operating temperatures for equipment that contains disk drives.

Google Study Examines Effects on PC Hard Drives

However, the study did find that hard drives suffer from "infant mortality," where newly-installed hard drives were more susceptible to temperature and other factors, even after a burn-in period. After a period of about six months, failure rates as a factor of temperature remained relatively stable until about three years, when they began to spike.

Moreover, in general, lower temperatures (about 20 degrees centigrade) produced higher failure rates. "The figure shows that failures do not increase when the average temperature increases," the paper noted. "In fact, there is a clear trend showing that lower temperatures are associated with higher failure rates. Only at very high temperatures is there a slight reversal of this trend."

Finally, the study looked at SMART predictive analysis, and concluded unfortunately that the algorithms did not generally serve as a model for future behavior. However, normal procedures, such as scan errors and reallocation checks, were a good indication that a drive would failure.

"The critical threshold analysis confirms what the charts visually imply: the critical threshold for scan errors is one. After the first scan error, drives are 39 times more likely to fail within 60 days than drives without scan errors," the paper said.

[noone_particular]

One thing I forgot to mention in my recommendation for longer HD life, use Linux. For the 24/7 people, Linux has far less HD activity during operation than Windows.

That doesn't surprise me at all. By the same token, 9X systems have a lot less hard drive activity than their NT counterparts. That probably contributes to the 9 year plus average lifespan the drives I'm using have had. Actually, we've most likely got this backwards. Hard drives have normally long long lifespans when used with good operating systems. It's the NT systems that make them die prematurely.

The bold print needs to be fixed, Heat= expansion and Cold= contraction, there's no way bearing/spindle fit is "a bit tighter" during startup.

Yes, I did word that backwards. Suffice it to say that the clearance does change as temperature changes. A slightly loose fit can also wear faster than a good fit. I'd bet that the looser fit also changes clearances at the disk, which would have effects of its own.

I've never seen the problem with running a PC 24/7. I have the monitor set to power down after 10 minutes of idle time but the PC is always ready. The PCs here get used at any odd time of the day or night, often for short time periods. Powering them down and restarting every time would result in a lot of restarts each day. A lot is made of the heat they give off, regarding it as waste. The heat is only wasted during the summer. The rest of the year, it's another minor source of heat. In this climate, that heat is welcome 2/3 of the time. It's energy the furnace didn't have to use.

[MrBrian]

Originally Posted by FastGame

One thing I forgot to mention in my recommendation for longer HD life, use Linux. For the 24/7 people, Linux has far less HD activity during operation than Windows.

The Google study addresses too. From the first link in my previous post:

Over work = early death?
A teenager might want you to believe that, but the Googlers found little correlation between disk workload and failure rates. Since most of us, including enterprise IT folks, have no idea how much “work” our drives do, utilization is a slippery concept. The authors defined it in terms of weekly average of read/write bandwidth per drive and adjusted for the fact that newer drives have more bandwidth than older drives.

"After the first year, the AFR of high utilization drives is at most moderately higher than that of low utilization drives. The three-year group in fact appears to have the opposite of the expected behavior, with low utilization drives having slightly higher failure rates than high ulization ones."

As the graph shows, infant mortality is much higher among high utilization drives. So shake that new drive out while it is still under warranty. And don’t worry about doing those daily backups to disk and other I/O intensive work.