July 25, 200619 yr Just ordered everything for a mass storage server yesterday, discovered unRaid today. Been browsing the forum for an hour, still have a couple of newbie questions... Having disks powering down is a very desireable feature, I guess it is implied that they (automatically) power back up whenever any process attempts to access their content? I don't understand how 1 parity drive can hold enough information to rebuild any 1 of the other 10 other drives, guess I'll wait for the Parity Explanation Page. Wonderful feature however! Only one/tenth the waste versus 1 to 1 mirroring, or is there redundant data on the sibling drives? Haven't seen any mention of staggered drive power-up sequence. If you have software control of the spin-state of a drive, seems like it would be easy to implement, if it isn't already. I went with ACard IDE controller cards. Any chance they will work in place of Promise TXn? Thanks for any answers you may have.
July 26, 200619 yr In this following post on AVS Forum I described how a single parity drive can be used in combination with any number of data drives and subsequently used to recreate any one of them. http://www.avsforum.com/avs-vb/showthread.php?p=6278295&post6278295 It is not that the parity drive stores a copy of all the data on all the other drives, it could not. That's physically impossible. It can store, in combination with all the other data drives, enough information to recreate any ONE missing drive in the event of a failure. Joe L.
July 26, 200619 yr Author Ah yes, I see. Took me about 2 seconds after seeing your AVForum example. Any missing bit can be deduced by examining the corresponding bit on each of the other drives, and then looking at the parity bit. I'll have to read it a little more to understand how this state of affairs can be maintained if update activity continues while the disabled drive is offline... Thanks!
Archived
This topic is now archived and is closed to further replies.