burnchain connection retry should periodically perform another DNS lookup

[wileyj]

if the burnchain host being used drops connection and comes back later with a different IP address, the node is unable to reconnect.

I think that periodically, or possibly on the 60s retry when reconnecting to the burnchain host, a DNS lookup should be performed again.
else, what i've seen happen is the chain essentially stalls and there's no real error in the logs (i found this out by doing the above: rebuilt a btc node and did not restart the stacks nodes attached to it. they all were stalled until restart).