Talk:Hard disk drive failure

no citations

This article has no citations for the many claims made. Is this normal and acceptable for this type of article?

No and no. I'm tagging it as needing citations. DBaK (talk) 09:43, 1 December 2009 (UTC)

driveservice.com link

Possible advertising, but more importantly some information there is dated. Would be an excellent reference if current, but some assertions at the link are not true today. For example, Fujitsu continues to make hard disks; several 300 GB SCSI drives are listed on eBay at the time of this writing. Marc W. Abel 20:17, 27 June 2007 (UTC)

Agree. Information is extremely outdated and subjective. Zzptichka (talk) 14:23, 26 October 2008 (UTC)

CD Failure

A rather bold statement saying that CD and DVD can fail in a drive. I have yet to find any evidence of this.

I was skeptical, but I added a site where some folks did testing on disks. 16x players run disks at over 25,000 rpm, and many disks shatter at 30,000. Surprising, but looks factual. DonPMitchell (talk) 16:45, 23 July 2009 (UTC)

I looked up a NIST study on optical media (CD/DVD) failure and added a few sentences about that. They recommend gold-layer media (e.g., Verbatim Gold Archival Grade and other brands, although I didn't mention brand names in the article). DonPMitchell (talk) 17:18, 23 July 2009 (UTC)

I'm confused why the DVD and CD-ROM pages don't mention the NIST study. Also, the DVD article seems to have different numbers and the CD-ROM page doesn't mention lifespan at all. This is all confusing to users like me who want to backup stuff but who research online and find a lot of bad reviews for HDDs quickly going bad all the time and seemingly random numbers for CD and DVD lifespans. It sounds like printed media like photographs last much longer. Thank you for adding what was available at the time. I hope more research can be done by organizations in the near future. Zeniff 21:35, 14 May 2010 (UTC) —Preceding unsigned comment added by ZeniffMartineau (talk • contribs)

Disk mortality bath tub curve citation

From "Disk failures in the real world: What does an MTTF of 1,000,000 hours mean to you?" ( http://www.usenix.org/events/fast07/tech/schroeder/schroeder_html/index.html ) section 4.2 (Age-dependent replacement rates):

"One aspect of disk failures that single-value metrics such as MTTF and AFR cannot capture is that in real life failure rates are not constant [5]. Failure rates of hardware products typically follow a "bathtub curve" with high failure rates at the beginning (infant mortality) and the end (wear-out) of the lifecycle."

144.32.48.87 (talk) 15:56, 27 September 2011 (UTC)

Predicting Archival Life of Removable Hard Disk Drives

Predicting Archival Life of Removable Hard Disk Drives

http://lockss.org/locksswiki/files/ISandT2008.pdf

Williams et al.

Abstract

A collaboration between the developer of a Removable Hard Disk Drive (RHDD) storage technology product and academics presents an accelerated life test of non-spinning powered down hard disks. The results are used to predict the reliability of removable disk cartridges stored in shirtsleeve environments as archival media. The contributions of this paper include an initial study of the reliability of data stored off-line on RHDDs, and a description of the industry standard accelerated life test process together with the methodology used to predict storage media reliability based on the results. Our study shows that removable hard disk drive technology has considerable potential as a reliable archival medium

There is ample evidence from the RAMAC Restoration that magnetic recordings will be there beyond 50 years. The harder issue is will there be any equipment in 50 years that can connect with and read a current HDD, or will it be, as is the case with the RAMAC, a scientific project to read the old hardware? Tom94022 (talk) 21:46, 2 January 2013 (UTC)

Tri-axis accelerometers

Perhaps the article could make mention of tri-axis accelerometers (eg Kionix KXPB5) used in some laptop drives. These are in addition to the normal shock sensors.

http://www.kionix.com/Product%20Sheets/KXPB5%20Product%20Brief.pdf
http://www.kionix.com/Product-Specs/KXPB5-2050%20Specifications%20Rev%203.pdf
http://www.kionix.com/Product-Specs/KXPB5-2353%20Specifications%20Rev%203.pdf
http://www.kionix.com/accelerometers/accelerometer-KXPB5.html

121.44.65.207 (talk) 01:58, 24 June 2013 (UTC)

The article all ready mentions accelerometers; most of the above links are broken and the Kionix site has nothing I can find about HDD applications, so I suggest we ignore this IP request. Tom94022 (talk) 18:18, 25 June 2013 (UTC)

Lack of specifics

I came here looking for some percentages, like what are the actual causes, for instance to compare how much of a role wear and tear play. I'm playing around with some ideas for customized RAID setups for fun, and I'm not sure to what extent I should be spinning drives down vs. scrubbing them as part of the algorithms. This article is pretty worthless overall. Imagine an article like this one called "Automobile mechanical failures": list all the parts, and yes, they can all fail. 198.179.82.133 (talk) 16:51, 18 October 2013 (UTC)

Moved above embedded rant into its own section. The info the IP wants is likely proprietary to the various subsystem manufacturers; lots of data are available in the references and See Also, so I'm not sure any action is necessary. Tom94022 (talk) 15:04, 20 October 2013 (UTC)

MY DRIVE WORKS ON ANOTHER COMPUTER, BUT NOT THIS ONE.

Hello, I am currently on a computer (Lenovo) on Windows 8. However, when I plug in the drive, the drive does not seem to be found, no matter how long I wait. When I plug it into another computer, it seems to work. Why is this?107.217.169.253 (talk) 19:54, 28 December 2013 (UTC)(A)

Bathtub

I suppose you can call this a bathub, although the dip in the "tub" is pretty small. 86.121.18.250 (talk) 20:23, 5 January 2014 (UTC)

It does look like one, but, from WD we have:

Does Hard Disk Drive Failure Rate Enter Steady-State After One Year?

It was found that the HDD hazard rate function doesn't follow the traditional bathtub curve, but manifests itself as an increasing-decreasing-stabilized pattern. Furthermore, its time-to-steady-state is not one year as historically assumed by the industry. As it turns out, the generalized Gamma distribution is a better candidate to characterize the HDD non-traditional bathtub curve than Weibull and other non-Weibull parametric models

IEEE Reliability and Maintainability Symposium, 2007. RAMS '07. 22-25 Jan. 2007

A bathtub curve is probably a reasonable approximation suitable for this article, but it will take some research to find a reliable sourcde. Tom94022 (talk) 22:15, 6 January 2014 (UTC)

Citations

http://citeseer.ist.psu.edu/viewdoc/download;jsessionid=99578A99113881E02C0514F2B5618152?doi=10.1.1.160.1618&rep=rep1&type=pdf has a table (Table 3 Disk Drive failure Mechanisms) talking about disk failure mechanisms and another section (3.1 Failure Modes) that talks about failure modes. http://www.eecs.berkeley.edu/Pubs/TechRpts/1999/CSD-99-1066.pdf also has (4.3.2. Disk Drive Cases) talking about fault types and their frequencies.

80.194.75.37 (talk) 11:55, 21 July 2014 (UTC)

Data recovery - "reading data from the platters directly"

The following statement in the "Data recovery" section is a myth:

"Specialised companies carry out data recovery, at significant cost, by opening the drives in a clean room and using appropriate equipment to read data from the platters directly."

Spin-stand Microscopy is not used by any data recovery shop simply because it would be astronomically expensive, even if it were possible on modern high density drives. The only procedures performed by DR shops are head replacements in those cases where one or more heads have failed, or platter transfers in cases of seized or worn spindle motors or bearings.

The following site has video tutorials which demonstrate the procedures involved:

https://hddsurgery.com/

106.69.141.20 (talk) 00:04, 10 April 2020 (UTC)

Agree, changed the article. Tom94022 (talk) 19:18, 10 April 2020 (UTC)