Home
| Calendar
| Mail Lists
| List Archives
| Desktop SIG
| Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings Linux Cafe | Meeting Notes | Blog | Linux Links | Bling | About BLU |
The attraction to raid 5 and 6 is that you don't loose nearly as much HDD space over raid 1. Say you have a 4u rack with 16 drives, you want to split them, you could get 8 drives and set them up with RAID 1, thus giving you a 50% HDD storage capability. If you want 2 LUNs on a raid 5 you have 14 drives at your disposal and if one drive dies you don't have a problem. With raid 6 you have 12 drives at your disposal and a failure of 2 per LUN, much much better then RAID 1 where if you loose 2 drives and they happen to be the same LUN you loose your data. ~Ben On Thu, Mar 6, 2008 at 8:54 PM, Matthew Gillen <[hidden email]> wrote: > Scott R. Ehrlich wrote: > > When I get a replacement disk and build the system from the ground up > > again, I could, conceivably, use hardware RAID 1 for the OS on two > > disks, and CentOS 5 64-bit's LVM for software RAID 5 (or maybe 1+0 if > > available) on the remaining for 4 disks, maybe 3 disks as active and the > > 4th as a hot spare? > > > > I've never had much faith in software raid, since it is not > > hardware-based, and there would be a performance hit, but in this case, > > it could be an option. > > From a purely logical standpoint (I don't have the inclination to do > actual > performance tests myself, since I don't have any RAID hardware, and the > software works fine for me), there's not much that hardware will gain you > over software when doing RAID-1. RAID-1 is all about bandwidth to the > drives, so as long as you have separate buses that can write to more than > one disk in parallel, that probably dwarfs any other issues. > > With RAID-5 the story is a little different, since the parity calculation > would actually chew up some CPU cycles. There, hardware-assist would > probably make a noticeable difference. Personally, I don't see the > attraction to RAID-5; it's more complexity for very marginal cost benefit > (which is probably negated completely by the cost of the RAID > controller...) > > I'd be curious to see some performance tests of a hardware RAID controller > vs. a couple SATA drives using SW RAID, for both the 1 and 5 flavors. > > Finally, one last point in favor of SW RAID: a bug in the implementation > can > be fixed easily. If your RAID-controller has a bug, you're fsck'd. > > Matt > > -- > This message has been scanned for viruses and > dangerous content by MailScanner, and is > believed to be clean. > > _______________________________________________ > Discuss mailing list > [hidden email] > http://lists.blu.org/mailman/listinfo/discuss >
BLU is a member of BostonUserGroups | |
We also thank MIT for the use of their facilities. |