I've a box with an ASUS M2N32-SLI Deluxe motherboard that should supposedly be supported by 'dmraid' which I am under the impression ships with Fedora and RHEL5.

http://www.nvnews.net/vbulletin/showthread.php?t=71735
http://nvidia.custhelp.com/cgi-bin/n...i=&p_topview=1

However at install time after entering the Installation Number, Anaconda hangs when it is time to detect partitions. The ideal target partition is a 2xSATA Raid0, and there is also a 2xSATA Raid1 block on the box.

Is it not possible to install onto Raid0, and if not should the installer just "hang"? Shouldn't it at least provide some information as to what is going on or notify there are no valid install targets? The entire Anaconda hanging (the box is still responsive but there is no useful info on any of the VTs) is a bit disheartening.

EDIT:
using Knoppix I can "see" the Raid setup as it should be using `dmraid` command.

In all of my experiences, Linux software RAID (using mdadm) is infinitely more stable than the "hardware" RAID built into the motherboards. As a bonus, it usually performs better, too!

Well I've been reading and it looks like I wouldn't be able to boot from RAID0 anyways (RAID1 ok) according to the RHEL5 Deployment Guide.

I really would rather just use linux software raid, but I have to convince the client who configured and purchased the machine (without mentioning it to me) that it really doesn't have hardware raid. He seems convinced the onboard controller is the way to go, so I've been attempting to make it work.

Looks like a no-go unless someone here has some definite insight... I'm on my way to convince him to either go linux software raid or purchase an extra drive for the linux installation and then we can use the onboard raid controller with dmraid and mount those seperately (for what I have no idea).