(2014-12-12, 20:19)MrCrispy Wrote: I've read on many places that Seagate reliability is less than Western Digital, but on the other hand the large scale studies in data centers by Google etc, who use consumer drives and not enterprise, say there is no real link between drive oem and reliability.
Who knows at this point.
Most quote blackblaze
What they fail to realize is
A) Blackblaze purchased refurbished drives in mass quantities of an already known drive that had an issue
B) Blackblaze builds it's own servers that run 24/7
C) Blackblaze has revised these server builds 4 times. During the time of the Seagate 1.5 TB issue (Also found in their 3TB drive it seems), they were on revision 2. They replaced that with Revision 3, specifically noting Vibrations in the chassis that were leading to larger than normal failures.
D)Blackblaze's conclusion of their article stated they'd use 4TB Seagate drives moving forward.
E) Blackblaze released an update 4 months ago. HGST had the lowest failure rates, the 1.5TB and 3TB drives from seagate still were bad(well no duh, refurbished drives of a drive with known failure issues will be bad), and 3TB drives from Western Digital had very high failure rates as well.
F) You aren't running your drives anything like how Blackblaze does so their reliability data is utterly irrelevant to 99.9% of consumers.
There isn't any long term data on the 4-6 TB Seagate drives to say they're bad but there is enough data on the 3TB WD drives to show that it may or may not have an issue (depending on what revision of their storage pods Blackblaze put them into).
Hence why I don't use their data in my purchasing decisions. People who do just show that they're capable of reading a chart, and are utterly incapable of actually looking at how the data was collected that went into a chart.