Hard Drive MTBF testing?

[Kakumba]

Sorry if this has been covered before, getting to grips with this forum atm.

So, my question is:

How do manfacturers test HDD's for Mean Time Before Failure? Esp for SCSI drives (they certainly dont leave them sitting there for 1.2 million hours, I'm fairly confident)

 

[Matthias99]

Originally posted by: Kakumba
Sorry if this has been covered before, getting to grips with this forum atm.

So, my question is:

How do manfacturers test HDD's for Mean Time Before Failure? Esp for SCSI drives (they certainly dont leave them sitting there for 1.2 million hours, I'm fairly confident)

Been asked (and answered, by me and probably others) before. Please search.

Generally speaking, it's a combination of accelerated and/or bulk lifespan testing, and statistical analysis of failure rates of parts that are common to existing drives. For drives that have been around for a while, 'real world' failure rates can also be analyzed.

 

[Philippine Mango]

Yea you might wanna try google or search around, basically they take a bunch of drives and run it for a set period of time and a few of those drives in that time period would fail...

 

[Kakumba]

Aye, I was gonna leave this thread alone, let it drop off, and do soem searching when I get home (at work right now, maybe I shouldnt be on here?)

 

[JimPhelpsMI]

Hi, With our engineers it was an educated guess based on the MTBFs of the components supplied by the individual parts manufacturers. Anyway yours may be the one that fails in the first hour or the one that runs twice as long as the MTBF. Kinda like an average time. It is said that if it makes it through the first month of use then it may last forever.
Also our guys always said "Electronics (Mechanics etc) is capable of 100,000 errors per millisecond. Luck, Jim