The impact of heavy use and high temperatures on hard disk drive failure may be overstated, says a report by three Google engineers. The report examined 100,000 commercial hard drives, ranging from 80GB to 400GB in capacity, used at Google since 2001.
“Our data indicate a much weaker correlation between utilisation levels and failures than previous work has suggested,” the authors noted.
“We expected to notice a very strong and consistent correlation between high utilisation and higher failure rates. However our results appear to paint a more complex picture. Only very young and very old age groups appear to show the expected behaviour.”
I like that they collect data about all their systems “every few minutes” and store it “essentially forever”. Sounds like my email filing system.
Funny about the timing of your report. Because Google has not been online for a long time today !
Can you access it ?
I hate it that BBC (and other traditional media companies) still haven’t learned how (or why it’s important) to link to the original site or document from which they sourced a story.
It’s like they don’t want you to know more about the subject you are reading. I guess they can’t really deal with the fact that they are no longer the big filters they once were.
Anyway, enough ranting. The original article (very cool, even though not as detailed as I expected initially) can be found here:
http://labs.google.com/papers/disk_failures.pdf
All of these threads, and this is the only one worth a damn.
Very interesting report, and thanks to #2 for the link. It seems that… if the report were used for a business case… it would be that a noticable number of drives fail off the bat… yet somehow escaped internal QA.
Is this something that can be improved upon by the hard drive makers, or is this just the nature of an analog device?
#1 — no trouble at all w/Google from the southern Rockies.
#3. The inverted bell curve of HD failures (early/late) is well known in the industry, so this aspect of Google’s report is not surprising. Some of the early failures can be traced to shipping, handling, and installation. If a drive is subjected to shock, esp. operational shock, head and media damage can create particulates that escape the internal filters, causing accelerated damage to the head and media.
#5 – No offense, but blaming the consumer eventually bites the seller in the ass.
It’s a device with moving parts, so some level of failure is expected, but what that level is could be improved upon.
1 – Yeah – I notice periodic “Google Outages” (TM pending 🙂 ) – I personally blame YouTube – its gotten worse since the acquisition, and YouTube must suck a lot of bandwidth – In completely unrelated news, one of the founders (sorry, I forget which, and can’t find the article) said that internet TV will never take off, because the internet in total doesn’t have the bandwidth to handle it – methinks the famed near-infinite bandwidth of Google is just slightly mythical.
#6. No offense taken. Users don’t typically install disk drives – OEMs do.
DNS has been acting up for days now. That’s the explanation for the outages.
Google been ok all day both at work and home (SC).
I think hard drive longevity, or lack thereof, is a crap shoot. I’ve got a housefull of the damned things that whizzed along for years until they became either too small or too slow for the job at hand. I’ve had one drive, a thirty GB Maxtor, that basically went tits up, and another, a Western Digital, that suffered a corrupt MBR, but I don’t hold that against the manufacturer, it’s just one of those things.
I have a 120 GB Maxtor external drive and that baby runs HOT whenever it is powered up.
… and I’ve kept it powered-up, 24-7, since 2002.
Can anyone beat that?
No need to “splain” it to the chimps.. joey