from my personal experience my veryold 4x 1tb WDreds still are working fine and are over 10 years by now (1 had checksum errors in freenas back in the day and got switched out)
my 6x 4tb WDreds are over 6 years by now and are still working flawlessly (:
This is the first time i bought a seagate drive in my life since it was recommended by backblaze i would be sad if it would not at least last as long
thatās the thing the manufactures tell another story than the community.
but since i dont trust 1st party sources that much i would like to have some verification.
An neither Backblaze nor other sources have some good studys regarding that topic
Anyways i more than halfed my power draw with switching from 6 drives to 1 (:
(and for anyone reading this in the future i have the old setup as cold storage and cloudsync - dont do 1 drive setups; )
An interesting white paper. Although the details are specific to a particular range of drives, I think most of details the intermediate stages of power saving, and that the 50%+ power savings for spindown (and 5-8sec spinup time) may be reasonably extrapolated as applying to most spinning hard drives.
iirc 2.5" drives are more power efficient by design / pure physical size since the motor has to spin less mass.
But i guess with a grain of salt the drives tested by back blaze fall into the category.
and relatively can see similar results?
The energy needed is probably a combination of angular inertia to get it to speed and then speed x surface area to cover friction with the gas surrounding the platters.
Because, for the same RPM, the outer 0.75" of a 3.25" drive moves faster than the inner 2.5", I suspect that the increase in energy is more than the ratio of surface areas - probably some sort of size ^ 3 factor i.e. perhaps 2.2x as much energy. But I havenāt attempted to use my 4 decades old calculus to work it out.
All 4-5 TB 2.5" HDDs are SMR, so basically unsuitable for ZFS.
Iām not sure about 2 TB drives, but at such low capacity thereās no reason not to go SSD.
