Who makes the most reliable hard drives?

[Mkruer 4]

Tech reports has pushed a article from an online backup provider Backblazehas that is currently using over 25,000 units of various sizes and discusses their finding on the most reliable HDD.

http://techreport.com/news/25940/hard-drive-reliability-study-names-name

the original information comes form

http://blog.backblaze.com/2014/01/21/what-hard-drive-should-i-buy/

[MRFS 20]

It's very interesting that "vibration" is mentioned as a key factor.

I remember a video that Allyn Malventano cites in his review

of WD's "Red" series:

http://www.pcper.com/reviews/Storage/Western-Digital-Red-3TB-SATA-SOHO-NAS-Drive-Full-Review/Specs-Testing-Methodology-an

WD's Reds have a spindle bearing mechanism that

works somewhat like this Centramatic Automatic Wheel Balancer:

The question that arises, then, is the amount of "damage"

that can be attributed to excessive vibration, and

how much "damage" can be obviated when any given

HDD is NOT subject to excessive vibration.

Nevertheless, the Hitachis in this next table

certainly score very well on "Annual Failure Rate":

http://blog.backblaze.com/2014/01/21/what-hard-drive-should-i-buy/

[begin quote]

Hitachi Drives

If the price were right, we would be buying nothing but Hitachi drives.

They have been rock solid, and have had a remarkably low failure rate.

Hitachi Drives Used by Backblaze

Model Size Number Average Annual
of Age in Failure
Drives Years Rate

Hitachi GST Deskstar 7K2000
(HDS722020ALA330) 2.0TB 4716 2.9 1.1%

Hitachi GST Deskstar 5K3000
(HDS5C3030ALA630) 3.0TB 4592 1.7 0.9%

Hitachi Deskstar 5K4000
(HDS5C4040ALE630) 4.0TB 2587 0.8 1.5%

Hitachi Deskstar 7K3000
(HDS723030ALA640) 3.0TB 1027 2.1 0.9%

Edited January 21, 2014 by MRFS

[MRFS 20]

I wrote to the CEO as follows today:

http://blog.backblaze.com/2014/01/21/what-hard-drive-should-i-buy/
http://www.backblaze.com/contact.html


Greetings Mr. Budman:

We don't manage anything close to the number of

HDDs you are purchasing and deploying.

Nevertheless, because that article ends with the question --

"What has your experience been?" --

please permit me to repeat some comments
that you may have already heard:

( 1 ) steady and reliable input power appear to be

very important to keep HDDs happy; and, this implies
that SHUTDOWNs should be scheduled and not sudden;

( 2 ) temperature appears to be a double-edged sword:

clearly, excessive heat is known to accelerate failure,
but one study by Google indicated that excessive cooling
may cause bearing lubrication to create more drag

on moving parts because of increased viscosity;

ask the manufacturers to recommend optimal operating temps;

( 3 ) cooling and cleaning also appear to interact:
when dust accumulates on the larger upper and
lower surfaces of 3.5" and 2.5" HDDs, that dust

interferes with the effectiveness of cooler air

as fans force cooler air to flow over those surfaces;

( 4 ) although this factor may not apply to your

larger storage server farms, we always resort
to short-stroked primary C: partitions on our
Windows workstations, and we defragment
on a regular basis;

( 5 ) along the same lines as short-stroked C: partitions,
we also move the Windows swap file "pagefile.sys"
to another short-stroked primary partition, and
we make sure that file is also perfectly contiguous,
using the excellent CONTIG freeware program;

( 6 ) although the extra cost may be prohibitive

for an operation as large as yours, we have found
that WDC's 5-year warranties come out ahead

in terms of price per warranty year; we have
very much appreciated the ability to replace

a RAID Edition WDC drive if it has failed
in years 4 or 5; clearly, with a 3-year warranty,

we would be unable to get a factory replacement;

( 7 ) we just bought a 3TB Toshiba and we're
very happy with it, particularly its low price;

although we haven't used it very long,

Newegg also offers their own extended warranty plan
that keeps the total cost below the comparable

prices of WDC's RAID Edition 5-year warranties.

Hope this helps: THANKS FOR PUBLISHING THAT STUDY!

--
Sincerely yours,
/s/ Paul A. Mitchell, B.A., M.S., Instructor,
Inventor and Systems Development Consultant

All Rights Reserved without Prejudice

[Lamb0 0]

All excellent tips and information.

I especially thank Backblaze for releasing the vendor specific information.

Though the Egg rating methodology might be questionable, the Backblaze results statistically corroborate many of hearsay reliability grumbles I've heard the last couple years. Where are the best choices for HGST pricewise for small quantities stateside where you can trust the shipping? I do NOT trust NewEgg with spinning rust! ProVantage? Amazon? TigerDirect?

It'll be awhile before I need the TB; but it appears I'll need the budget boosted a bit before I expand beyond the 74GB Z2 test (3 Raptors) siqnificantly. That's OK, I'd rather make a much larger proportion of my mistakes during the learning process with a much smaller collection of filesystems!

A simple RAID1 OTOH...

[MRFS 20]

Toshiba must have been reading complaints about Newegg's OEM packing problems.

One user recommended the retail boxed version, so we tried one and

indeed the styrofoam insulation was very generous:

http://www.newegg.com/Product/Product.aspx?Item=N82E16822149396&Tpk=N82E16822149396

[jtsn 5]

Although my sample size is much smaller (about 30 drives from all four vendors), the report reflects my reliability experience quite well. Seems, that Seagate is the new IBM Deathstar.

[cppguru 0]

This is a can of worms to be honest. I have seagates with 0 failures for about last 9 years. I think it's all about the environment the drive is operating under. For instance, is there construction going on near your house?

The problem really is shock While the drive is spinning. I have personally dropped hard disks multiple times (when not spinning) and it just seems to power back up as if nothing happened. I have tried this from about 5 feet height with multiple drives (Hitachi, Seagate, WD). Looking at the white sheet from Seagate does support this as well if you look at their max Gs for the two cases (parked, operating).

If you are THAT concerned about your data, run two hard drives in Raid 1 and be done with it.

Edited February 10, 2014 by cppguru

[jtsn 5]

I think it's all about the environment the drive is operating under.

For my own sample the environment is the same for all drives. The Seagates were the drives I had most issues with, while Hitachi are the most solid. And WD and Samsung in between with just being average. I just refrained from early conclusions, because I don't have enough empirical data. But I can confirm Backblaze's impressions.

For instance, is there construction going on near your house?

If a drive fails, because there is a construction site nearby, it's broken by design. Thankfully working HDDs don't do that.

If you are THAT concerned about your data, run two hard drives in Raid 1 and be done with it.

I have a working backup strategy preventing data loss (which isn't the purpose of RAID1). But failing drives are a hassle and don't deliver my money's worth. So I'm still interested in avoiding lemons.