TheRegister about the Triple Parity in ZFS
The story about the Triple Parity finally reached the Register today - Sun tripling RAID protection. I won´t comment the articles style (“Hey, that´s old … and all other will have it anyway soon”). I would be astonished, if they had written “Hey, Sun … RAID-TP in a Tier-1 product … well done”. But i want to comment this article:
That has to be logically true but, if the use of 3.5-inch disks switches over to 2.5-inch drives then that would reduce failed disk rebuild times. It would also likely increase the number of drives in an array, putting us back, roughly speaking, at square one.
I don´t think that this really true, at least not at this extend. You should consider that the industry reached the 1 TB a few days ago with the Western Digital WD10TEVT. In the 3.5” world the industry is able to put 2 TB in one drive. Now just a little thought game: The Spinpoint M6 demonstrated, that´s possible to put 3 platters into 9.5mm. Lets assume that 1 milimeter on each side is used for casing and electronics. That´s an per-platter height of 2.5 mm. The maximum height of an 2.5” is specified with 15 mm. Let´s assume 1 mm for casing and electronics again. 13 mm for disks you would be able to put 5 platters into a 15mm height 2.5” casing. Switch back to this new 1 TB 2.5” drive. It provides 333 Gigabyte per platter. Such a fat small disk would provide you with 1660 Gigabyte on 2.5” and that´s not that far away from the 2 TB we have in 3.5”. Of course you would not use such a disk in a notebook (to large, to fragile for mobile, too loud), but we talk about disks for usage in the controlled environment of storage chassis. So, dear Chris Mellor, switching from 3.5” to 2.5” isn´t a jump back to square one. At the maximum it´s a jump one square back …