I happened to see this discussion in the archives, and thought I'd
throw my two-cents in.

	I'm running a 3-disk raid5 system on my 1.6.1-stable box with 3 IDE
disks across three IDE buses, with a hot spare in place, and it seems as
stable as it could be.  I've got softdep enabled on all filesystems, and,
as long as I don't swap to the raid-5 set, it runs for weeks and weeks.  If
I start exercising my swap space heavily, by running an application which
sucks a lot of memory and ten drops it when it exits, I can induce the
system to hang, but only after the application has been running for
multiple days.
	Previous to this, I ran a system with a raid-5 system on it which had
15 disks in its raid set.  This system rand fine for two years, until all
of the IBM DethStars (75GB IBM DeskStars) died at once and we abandoned
that configuration.  Both systems are run pretty heavily, and both
experience a lot of network traffic and a lot of disk read/writing.
	I wonder if there is something particular about your configuration
wich causes raidframe to behave badly?
-Brian