On 05/03 05:50 , John Hudak wrote:
> Interesting....what tests in bonnie++ did you run?
> DB access to a single file?
> Create/write/delete small files?
> Others?

Probably just create/read/rewrite many small files. 

> Just curious, so you beat the disk senseless with bonnie++ for a day, and if
> the drive does not fail, when put into service, how long does it last (based
> on your experience)?

The ones that survived haven't died yet. :)
I just use them for occasional archival storage tho; so they hardly ever run
now.

> How was this one day duration arrived?

based on when I got bored. :)
I had some disks die after 9 or 12 hours. If it lasted longer than that, it
was likely to last the 2 days or so that I ran the longest tests.

> (The implication is that if the drive survived the test for one day, you can
> rule out infant mortality, and expect the disk to last to meet the mfg
> MTTF?)

pretty much. they mostly seemed to die within 9-12 hours. 
It should be noted that my sample size was not statistically significant.
Perhaps 7 different disks in 5-6 different enclosures. Don't remember
anymore. Was 4-5 years ago.

> I have read several anecdotal articles that concluded USB and FW disks in
> their own enclosure have a higher probability of infant
> mortality....unfortunately they weren't tested in a scientific manner....

My tests were far from scientific, and the plural of anecdote is not data.

-- 
Carl Soderstrom
Systems Administrator
Real-Time Enterprises
www.real-time.com

------------------------------------------------------------------------------
_______________________________________________
BackupPC-users mailing list
BackupPC-users AT lists.sourceforge DOT net
List:    https://lists.sourceforge.net/lists/listinfo/backuppc-users
Wiki:    http://backuppc.wiki.sourceforge.net
Project: http://backuppc.sourceforge.net/