[fw] Backblaze lists most reliable hard drives based on its massivecloud study

  • Thread starter Thread starter Mr. Man-wai Chang
  • Start date Start date
M

Mr. Man-wai Chang

The results from three years of use were revealing: Western Digital's
drives lasted an average of 2.5 years, while Hitachi's and Seagate's
lasted 2 and 1.4 years, respectively. Even so, some of the individual
Hitachi models topped the reliability charts.

"Hitachi does really well. There is an initial die-off of Western
Digital drives, and then they are nice and stable. The Seagate drives
start strong, but die off at a consistently higher rate, with a burst of
deaths near the 20-month mark," Backblaze wrote in its official blog.
"Having said that, you'll notice that even after 3 years, by far most of
the drives are still operating."

http://www.computerworld.com/s/arti..._based_on_its_massive_cloud_study?source=cwfb



--
@~@ Remain silent. Nothing from soldiers and magicians is real!
/ v \ Simplicity is Beauty! May the Force and farces be with you!
/( _ )\ (Fedora 19 i686) Linux 3.12.8-200.fc19.i686
^ ^ 19:06:02 up 2:56 0 users load average: 0.01 0.04 0.05
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
Mr. Man-wai Chang said:
The results from three years of use were revealing: Western Digital's
drives lasted an average of 2.5 years, while Hitachi's and Seagate's
lasted 2 and 1.4 years, respectively. Even so, some of the individual
Hitachi models topped the reliability charts.

"Hitachi does really well. There is an initial die-off of Western
Digital drives, and then they are nice and stable. The Seagate drives
start strong, but die off at a consistently higher rate, with a burst of
deaths near the 20-month mark," Backblaze wrote in its official blog.
"Having said that, you'll notice that even after 3 years, by far most of
the drives are still operating."

http://www.computerworld.com/s/arti..._based_on_its_massive_cloud_study?source=cwfb

Yeah, we're all sitting back with our bag of popcorn,
waiting for the Seagate press release :-)

Paul
 
The results from three years of use were revealing: Western Digital's
drives lasted an average of 2.5 years, while Hitachi's and Seagate's
lasted 2 and 1.4 years, respectively. Even so, some of the individual
Hitachi models topped the reliability charts.

"Hitachi does really well. There is an initial die-off of Western
Digital drives, and then they are nice and stable. The Seagate drives
start strong, but die off at a consistently higher rate, with a burst of
deaths near the 20-month mark," Backblaze wrote in its official blog.
"Having said that, you'll notice that even after 3 years, by far most of
the drives are still operating."

I went sour on WD from CompUSA Sun morn while supply lasts sales.
Seems they had clocks to set them to break at precisely near a year,
including warrantee replacements. Funny after 10, 15 years that the
game still hasn't changed, other than SSDs as boot option only.
 
Over the years, and with several dozen HHDs, I've had better luck with WD (blacks and
reds only since they've been color coding) than with the others. Hitachi has been
responsible for the worst experiences.

Hitachi took over IBM. I didn't think much of them, IBM, back when
they started outsourcing to oddball places such as Malaya, Turkey,
Pakistan, whatever...2- to 6-hundred MEG capacity circa;- switching to
WD, I don't know, pre-1T drives, when WD had a defense contract and
touted themselves for supplying the US Navy's needs;- after that,
moving up to T-class, Seagates, whenever I can got them, FUJI and
whatever else then that has garnered some decent reviews.

Most of those, a few "green class" drives, couple blacks, sit off to
the side for three storage docking stations (some of which work lower
and won't hit 2T) -- all being either 1.5 or 2T drives.

Those CompUSA WD's, short-term burn-out sales items, really burned my
butt, though way back when, it's so probably time now I went back with
half-a-mind better not to buy on impulse sales. I stocked up
pre-typhoon, no more $59 2T drives, although I could handle a 3T for
something less than a bill, WD withstanding, do-able deals now if for
nothing more than a challenge of "software" partitioning it out on
some transitional scheme with older, 1st-gen dualcores MB BIOS-es --
XP/SP1 complain or no W7 dependencies. Heh - Got to hold on to those
things or they'll have you somewhere, in shit-for-shineolaland,
floating off on some cloud service.
 
Paul said:
Yeah, we're all sitting back with our bag of popcorn, waiting for the
Seagate press release :-)

Western Digital acquired Hitachi in 2011.
Seagate acquired Samsung also back in 2011.

Maybe they're still using different manufacturing sites despite the
acquisitions.
 
Western Digital acquired Hitachi in 2011.

Hitachi was later surrendered to Toshiba as part of the deal!

--
@~@ Remain silent. Nothing from soldiers and magicians is real!
/ v \ Simplicity is Beauty! May the Force and farces be with you!
/( _ )\ (Fedora 19 i686) Linux 3.12.8-200.fc19.i686
^ ^ 23:39:02 up 2 days 7:29 0 users load average: 0.00 0.01 0.05
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa
 
The results from three years of use were revealing: Western Digital's
drives lasted an average of 2.5 years, while Hitachi's and Seagate's
lasted 2 and 1.4 years, respectively.


So you should replace half of your drives ever darn year?
 
So you should replace half of your drives ever darn year?

I didn't read it, although I'd hazard it's a case study in a 24/7
abuse scenario, likely networked and streaming for continuous access
and non-stop rewrites. I've drives here, Seagate 200G capacity, that
are going on 5-7 years continuous runtimes, 24/7, just without all
that abusive activity. Quite the contrary, I'm a bit of a gearfreak
optimizer, been messing with caches and compression schemes since
Day1.

Well...I did have one, so far, that's went belly up, I believe as a
result of a failed, undetected front-case fan;- In any event, before
this summer, I'm taking a Dremel tool, 1/8" shanks to cut out and
customize the front plastic facade for all my cases, (including an
expensive Antec), so I can be visually stimulated with the exactitude
of what, precisely, the hell is going on. Zero drive-failure
tolerance and a new leaf for paging through my newest book.

(In any event, last study I looked at, I doubt would be radically
unrevealing, on a concensus on HD-to-brand failure skews are there
really aren't any: There are both plenty of bad and good drives in any
batch/make, so just buy what you got and bring it along for today's
lunch on Potluck St.)
 
Davej said:
So you should replace half of your drives ever darn year?

You replace them as the SMART statistics, or the drive
performance, dictates.

In my case, I replaced the same drive on a yearly basis,
twice, and eventually got a clue and changed brands. We'll
see how well that works out, in a year or two.

In other words, it's adaptive behavior, until you find
something you can trust.

The nice thing about the drive "failures" here, is you
can use SMART to get some idea whether the drive is in trouble
or not. Whereas, I've had a few Maxtor drives which just failed
in the period of a day or so (before a backup could be made).
Modern drives can still fail due to firmware issues, but I
detect that possibility, by reading forums where outraged
users discuss the latest firmware issue (like forums.seagate.com).
Example of a search term I might use, before purchase.

site:forums.seagate.com st32550n firmware

That's what you'd enter, if you were planning on buying a ST32550n
hard drive from Seagate. Or maybe you'd search for "st32550n problem"
or "st32550n spin down" or the like.

If you want to view the SMART parameters, keep a copy of HDTune
handy. That's what I use. The free version has enough capabilities,
to keep track of your disk. The purchased version includes read/write
capabilities. If you like the tool and want to reward the author.

http://www.hdtune.com/files/hdtune_255.exe

The ones in red boxes should be zero. The scale is not linear,
and the value "zero" means "we haven't passed the threshold yet".
The drive can spare out a *lot* of sectors, before problems
show up in the red boxes. And for me at least, I hardly ever
find a Current Pending Sectors problem. Which is baffling,
unless the parameter just doesn't work right.

http://imageshack.us/a/img10/2134/cffn.gif

Some failing drives, you use the "benchmark" tab and benchmark
the drive. The occasional downward spike in the graph, is OK.
It is the nature of these benchmark curves, that they're interrupted
by other activities on the computer. For example, when the benchmarker
"crawls over" the pagefile, the computer may not behave quite normally
there. And I don't know exactly why. In any case, if the area where
your OS sits, the benchmark graph has a pronounced and sustained
"bad spot", you might use that as a replacement criterion. On my
last drive failure, that's why the drive was replaced. The two
zero values in the red boxes, did not betray that failure was
a possibility. SMART is not a guarantee of problem detection,
merely an additional tool to keep an eye on.

In the past, we wouldn't nave looked at SMART, and the drive would
just seem to die one morning, taking all the data with it. That's
how it used to work. Firmware issues can also cause abrupt failures,
but at least for one of the firmware-based failures (data structure
corruption), recovery of the drive was actually possible without
disassembling the drive entirely. The drive in that case, had a serial
interface with TTL logic levels, and a cell phone USB serial adapter
could be used to "talk to" the drive, and give it commands. The
command set is not documented, and some rocket scientist figured
out what voodoo you had to type to reset the drive. Expect that
very few drive problems, will be fixed that way. Without
documentation of the command set, we haven't a clue what
capabilities are in there.

Paul
 
Over the years, and with several dozen HHDs, I've had better luck with WD (blacks and
reds only since they've been color coding) than with the others. Hitachi has been
responsible for the worst experiences.

I've had good experience with WD, good with Hitachi (before they were
bought at least, I haven't used any since) and terrible with Seagate.
 
| On Wed, 22 Jan 2014 19:10:09 +0800, "Mr. Man-wai Chang"
|
| >
| >The results from three years of use were revealing: Western Digital's
| >drives lasted an average of 2.5 years, while Hitachi's and Seagate's
| >lasted 2 and 1.4 years, respectively. Even so, some of the individual
| >Hitachi models topped the reliability charts.
| >
| >"Hitachi does really well. There is an initial die-off of Western
| >Digital drives, and then they are nice and stable. The Seagate drives
| >start strong, but die off at a consistently higher rate, with a burst of
| >deaths near the 20-month mark," Backblaze wrote in its official blog.
| >"Having said that, you'll notice that even after 3 years, by far most of
| >the drives are still operating."
|
| I went sour on WD from CompUSA Sun morn while supply lasts sales.
| Seems they had clocks to set them to break at precisely near a year,
| including warrantee replacements. Funny after 10, 15 years that the
| game still hasn't changed, other than SSDs as boot option only.

Over the years, and with several dozen HHDs, I've had better luck with WD (blacks and
reds only since they've been color coding) than with the others. Hitachi has been
responsible for the worst experiences.

Larc

My experience seems to kind of agree with that of the study. I've had
drives in service on the computer, as well as set up as external eSATA
drives connected to HD DVR (3 of them) which means continuous service
recording. I've had both Seagate and WD failures (1 ea), but no
Hitachi failures. Incidentally, both the Seagate and WD failures were
while under warranty (5 years) and both were replaced. These were all
1 TB drives. The newer drives don't come with a 5 year warranty, so
I'm doubtful they will survive long enough to fail under warranty.

I have several 500 GB Seagate drives that were in service in RAID on
the computer, and they now have enough bad sectors reported that I no
longer use them in RAID..... but they still work and I sometimes use
them when I need some temporary storage.

My current plans are to simply buy the cheapest, on sale drive
regardless of brand.... and make frequent backups.
 
Back
Top