playing with software RAID

[Anand Buddhdev <arb anand org>]

I've just installed redhat 6.2 onto a Sun E250 server I have. It has
2 SCSI disks in it. Using the graphical installer, I setup software
RAID1 on this system, and created 4 devices, md0-md4, for the different
partitions, ie.  mirrors of each other. This is all fine so far, but I
am quite new to RAID.  I do not want to wait for a disaster on one of
the disks, and I want to be ready for when there is a problem. Since
the server is not yet in production, I want to play around with this
configuration.

Can someone suggest things I can do to familiarise myself with possible
scenarios? I was thinking of shutting down the box, and taking one disk
out, to see how the system behaves when restarted (to simulate a failed
disk). Is this a good test? And if a disk really does fail, how do I
tell? And if I then insert a brand new disk in place of a failed one,
how do I set it up to be a part of the RAID configuration? Will the RAID
system automatically detect a new disk and rebuild the data on it?

How can I add a third disk and also make it part of this RAID set? Sorry
to ask such basic questions, but I'd really appreciate some help and
pointers.

Thanks,

-- 
Anand Buddhdev