James Chacon wrote:
> I guess if you're only talking older hardware, sure. But this is the exact
> setup we've used to run stand alone sun's (ss20's, u1's, u2's, netra t1's, etc)
> as service machines (DNS, DHCP, etc) at work for a long long time.
> 
> With hundreds and hundreds of machines we experience single disk outages. Not
> once in recent memory (last 6-7 years) have I seen spindle lockup on both
> drives after a power cycle...But, with everything mirrored (including swap)
> one can survive a single disk outage and fix it during a more convenient time.
> 
> James

Hasn't happened to me yet, knock on wood, but some people have reported
more than one drive breaking down in RAID arrays.