Windows Home Server doesn't need raid. It does a sort of storage pool across all hard drives you add to it, and just the data you want will be replicated on more than one. They don't have to be identical drives, or sizes, or even busses (EIDE and SATA are fine together). It doesn't force you to do the same thing with all data, so if you have stuff that is easily replaced, it doesn't replicate it. It also recognizes identical files and only stores one copy of them, so if you're backing up 15 machines, you'll only have one copy of the OS files. Very cool stuff. It doesn't do a whole lot, but the few things it does it appears to do very well.
Sweet. Unique file identification is awesome, and it saves a LOT of space on backup media. We have an EMC box that checks file hashes when performing backups, so it only ever stores one copy of any given file no matter how many times it appears in the scheduling. It makes me happy.
The issue in my post came from developers using a storage device with known stability issues as a repository for actual important data, rather than using it to test file transfer speeds as it was intended for.
Re: Pennsic is coming!therejectMay 23 2007, 21:21:59 UTC
marasca has made the foolish mistake of sponsoring me this year, so I'll be pulling pleasureboy duties with the Regnesfolke-Myrkfaelinn. I shall bring offerings of meat.
Comments 7
Reply
Reply
Reply
The issue in my post came from developers using a storage device with known stability issues as a repository for actual important data, rather than using it to test file transfer speeds as it was intended for.
Reply
So get with the running around nekkid.
Reply
Reply
Reply
Leave a comment