Well, what's more interesting is a failover system when you poll several
sources of the same channels (maybe that's how yum already works, from
your description it doesn't look like this)
it is in fact:
yum has a failovermethod which lets it roundrobin or priority failover
b/t repos of the same packages example:
[somerepo]
name=some guy's repo
baseurl=url://somewhere/somepath
url://somewhereelse/somepath
url://somewhereelse/someotherpath
failovermethod=roundrobin
-sv