Unfortunately, I don't have any hard numbers, but my experience comes from the Desktop environment and the "server grade" drives seemed to do a little bit better than when I used consumer grade drives. Perhaps that handle power cycling better. I do wish I had kept really good notes though since I had at least 10,000 drives go though my hands, and it would be interesting. Brian On 05/19/2010 06:58 PM, Joseph Sinclair wrote: > A statistical analysis was published on a variety of disk drive lines a few years ago with a sample size of well over 100,000 drives[1]. > Other studies have shown negligible actual reliability differences between server and consumer lines across large populations. > Quality on consumer drives tends to be more variable, but no lower overall. > > I've seen all of the major brands work and fail at similar rates across large numbers of drives. Sometimes a batch will be bad, and age-in-store has a huge impact on reliability. > > Overall, run load as heavy as your main application will do for the first 3 months, as many drive failures seem to happen then (and it'll be in warranty). > After that, the next spike in failure happens at around 2 years and failure rates remain high thereafter. > > MTBF numbers seem to be largely irrelevant. > > Interestingly, SMART data seem to also be largely useless as well. > > > [1] http://labs.google.com/papers/disk_failures.pdf > > Brian Cluff wrote: >> I'll third on western digital, seagates are also pretty good, but >> whatever you do don't even start to consider Maxtor. They seem to have >> a near 100% failure rate. It's not if they are going to fail, but when. >> >> Also don't go with any hard drive manufacturers budget line or low end >> drive, try and go for drives that are made for server. They aren't all >> that much more, but seem to have much lower failure rates. >> >> Brian Cluff >> >> On 05/19/2010 03:14 PM, Shawn Badger wrote: >>> +2 on Bonnie++ >>> >>> I also tend to like the Western digital line for IDE/SATA drives. I have >>> had a few bad ones (out of hundreds), but they tend to run well for me. >>> >>> >>> >>> On Wed, May 19, 2010 at 3:09 PM, Alex Dean>> > wrote: >>> >>> >>> On May 19, 2010, at 5:07 PM, Nadim Hoque wrote: >>> >>> Hey, >>> >>> I am buying a new hard drive for my computer and I was wondering >>> what are some good hard drive stress test and how long should I >>> let it run for. I also do not mind what platform (windows, mac, >>> or linux) it will run on. Speaking of hard drives, which brand >>> do u guys recommend? >>> >>> >>> Bonnie++ is a disk benchmarking tool, but you can use it for stress >>> testing as well. Pretty easy to compile on Linux. I'm not sure if >>> it works on OSX or windows. >>> >>> alex >>> >>> --------------------------------------------------- >>> PLUG-discuss mailing list - PLUG-discuss@lists.plug.phoenix.az.us >>> >>> To subscribe, unsubscribe, or to change your mail settings: >>> http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss >>> >>> >>> >>> >>> --------------------------------------------------- >>> PLUG-discuss mailing list - PLUG-discuss@lists.plug.phoenix.az.us >>> To subscribe, unsubscribe, or to change your mail settings: >>> http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss >> >> --------------------------------------------------- >> PLUG-discuss mailing list - PLUG-discuss@lists.plug.phoenix.az.us >> To subscribe, unsubscribe, or to change your mail settings: >> http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss >> > > > > --------------------------------------------------- > PLUG-discuss mailing list - PLUG-discuss@lists.plug.phoenix.az.us > To subscribe, unsubscribe, or to change your mail settings: > http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss --------------------------------------------------- PLUG-discuss mailing list - PLUG-discuss@lists.plug.phoenix.az.us To subscribe, unsubscribe, or to change your mail settings: http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss