🚀 go-pugleaf

Thank you to all who responded.  Unfortunately, the hard lesson was taught
to us again.  The problems were attributed mostly to T-1 line errors!
With our Ethernet load on a weak T-1, things worked ok.  When the Token-
ring traffic was added (a substantial amount), the T-1 errors caused a
massive number of packets to be dropped.  Not all traffic died, but the
"new" Token-ring traffic was unusable, therefore, we concentrated too much
on this.  I am still pondering why this was not obvious, but at least what
we wanted to accomplish is much nearer, now.

(The more things change, the more they stay the same.)