Re:The Shingled magnetic recording NAS debacle
by phire on Tuesday May 26, 2020 @05:31PM
(#23316647)
It's quite clear that WD never properly tested these SMR drives in actual 3-8 bay NAS workloads. Certainly not a rebuild.
Based on their initial response to this scandal, it seems at some point they defined the "small NAS" Red spec as 180 TB/year read/written per year.
Then they later certified the new SMR drives as meeting this 180 TB/year spec, probably by testing they were good for 500gb/day.
They completely ignored the fact that larger home NAS installs don't use their 180 TB/year spec at a uniform rate, but instead tear though several TB doing a rebuild then sit near idle for most of the rest of the year.
Based on their initial response to this scandal, it seems at some point they defined the "small NAS" Red spec as 180 TB/year read/written per year.
Then they later certified the new SMR drives as meeting this 180 TB/year spec, probably by testing they were good for 500gb/day.
They completely ignored the fact that larger home NAS installs don't use their 180 TB/year spec at a uniform rate, but instead tear though several TB doing a rebuild then sit near idle for most of the rest of the year.