News for hackers. Stuff that matters.
Re:The final step for huge-page swapping by Dylan16807 on Wednesday July 04, 2018 @11:14PM (#17460632)
> Density does matter. And 12 individual disk will have a high failure rate than one disk because there is 12x electronics, 12x motors, 12x seals.

If you absolutely have to store 12TB of data and not spend another dollar, more drives are less reliable. In most situations having more drives makes it a lot easier to protect your data, because you can use parity.

> And you would need more raid cpu power to coordinate 12x disks.

A very small CPU cost; the drives are only doing 100 IOPS. And in return you get 4x-10x the throughput.

Re:The final step for huge-page swapping by mmt on Wednesday July 04, 2018 @11:29PM (#17460685)
> If you absolutely have to store 12TB of data and not spend another dollar, more drives are less reliable. In most situations having more drives makes it a lot easier to protect your data, because you can use parity.

I think what we're talking about is aggregates, so think about 120 1TB drives versus 10 12TB drives, if that's helpful.

If the failure rate per drive is comparable, then the failure rate per TB is 12 times higher for the the smaller drives. Arranging them in a RAID doesn't change this calculation.

> A very small CPU cost; the drives are only doing 100 IOPS. And in return you get 4x-10x the throughput.

I did stipulate to ignoring any performance difference, so we can assume that those 12 drives are concatenated rather than RAIDed, which I believe is an immesurable amount of difference in CPU load.