Are 3.5" HDDs more reliable than 2.5" HDDs by design?

[Pc Fixed Right]

what about 2.5" SSD drives they are pretty reliable no moving parts.

[Rob_NNCC]

http://www.datacenterknowledge.com/a...more-reliable/

http://www.google.com/?source=search...=1920&bih=1056

Short story, don't buy OCZ if you want it to last longer than a mechanical HDD.

[Rodrick]

Originally Posted by Rob_NNCC

http://static.googleusercontent.com/...k_failures.pdf

Heat does not kill HDDs, unless it is excessive heat like in the trunk of a hot car on a summer day.

The relevant quote from that report:

• Contrary to previously reported results, we found very little correlation between failure rates and either elevated temperature or activity levels.

[YeOldeStonecat]

Originally Posted by Pc Fixed Right

what about 2.5" SSD drives they are pretty reliable no moving parts.

I've notice people that work with them a lot complaining that they're failing at even higher rates than standard spindle drives.

One of the vendors we use for our edge firewall appliances....started using SSDs instead of Seagate Momentus drives. Figured more reliable right? Nope...they experienced over 20% failure rates within a few months....I have one of the units that failed at a client.

[PCX]

Originally Posted by Rob_NNCC

http://static.googleusercontent.com/...k_failures.pdf

Heat does not kill HDDs, unless it is excessive heat like in the trunk of a hot car on a summer day.

I use one of the Dell USFF PCs a SX280 (another SX280 is a hot standby) and besides from the cap issue, it has been fine. No issues with HDDs failing. I have been using them as HTPC and as our Admin PC in the shop so they have 24/7 usage for about 4 years under my ownership. (I got both due to popped caps.)

As far as reliability goes, given the same motion and heat/cool cycles, a laptop drive can handle extreme environments better.

If they where both un-moving and kept at 30*c, I would say the desktop drive due to larger parts and not an emphasis on small light weight designs, but then again it would matter quite a bit on the make and model.

Originally Posted by Rodrick

The relevant quote from that report:

As an observation, I agree that older drives are most definitely more likely to fail due to heat than newer drives. I have seen a definite trend in laptops that run extremely hot and HDD failures that follow. I do not agree that 9 months is not enough time to reach any real conclusion. Common sense and good practice still says that you should keep your drives cool.

The HP's with the AMD / Nvidia chipsets come to mind when correlating heat and failing hard drives.

[YeOldeStonecat]

A long long time in the IT industry has led me to agree with "heat and hard drives do not mix well".

I have noticed an unusually high correlation with the following scenarios which reinforce that.

*Desktops layered with dust bunny blankets inside..that smother things, and clog up the air flow to the hard drive(s)...lots of drive failures

*I've seen servers...built, where hard drives are right on top of each other...little to no air flow over them..higher rate of failure. One particular example that comes to mind...some guy built an HP Proliant tower server for an aircraft manufacturing companies Lotus Notes mail server. (yuck). He stacked the 3 drives right on top of each other in the 3.5" drive bay. Instead of mounting them in the 5.25" caddies and spacing them out in the 5.25" drive bays with plenty of air flow. Sure enough...2 drives failed in that rig.

*Server rooms where machines run 24x7..but are not kept to proper server room air temp standards..higher rates of failure. I've seen server rooms where the AC went on the fritz in the summer...over a weekend..come in Monday morning and it's over 90 degrees in there and some lights are flashing red to replace failed drives in the array.

*My best clients where I've been involved in planning their server room...I have it cooled to 65* on the nose. Have very..very little drive failure issues.

*I have a colleague that works out in Arizona...desert area. Wicked high temps. Deals with lots of drive failures in servers and workstations due to the brutal conditions.

[PCX]

Originally Posted by YeOldeStonecat

I long long time in the IT industry has led me to agree with "heat and hard drives do not mix well".

I have noticed an unusually high correlation with the following scenarios which reinforce that.

*Desktops layered with dust bunny blankets inside..that smother things, and clog up the air flow to the hard drive(s)...lots of drive failures

*I've seen servers...built, where hard drives are right on top of each other...little to no air flow over them..higher rate of failure. One particular example that comes to mind...some guy built an HP Proliant tower server for an aircraft manufacturing companies Lotus Notes mail server. (yuck). He stacked the 3 drives right on top of each other in the 3.5" drive bay. Instead of mounting them in the 5.25" caddies and spacing them out in the 5.25" drive bays with plenty of air flow. Sure enough...2 drives failed in that rig.

*Server rooms where machines run 24x7..but are not kept to proper server room air temp standards..higher rates of failure. I've seen server rooms where the AC went on the fritz in the summer...over a weekend..come in Monday morning and it's over 90 degrees in there and some lights are flashing red to replace failed drives in the array.

*My best clients where I've been involved in planning their server room...I have it cooled to 65* on the nose. Have very..very little drive failure issues.

*I have a colleague that works out in Arizona...desert area. Wicked high temps. Deals with lots of drive failures in servers and workstations due to the brutal conditions.

I am sure you have run in the same failures as I have then. Most of the failures I see due to overheating are read errors.