QoS/CoS interest

Scott Bradner sob at newdev.harvard.edu
Thu May 22 13:30:03 UTC 1997


Bill,

--
However, I always alike to be a bit contrarian and point out that QoS or
Multicast may never be needed because of the explosive growth of fiber
bandwidth.  I believe, in the future, it will be a lot easier and cheaper to
deploy bandwidth rather than manage complex router/switch technology to
support QoS/CoS.
--

I think this is a myth, at least for now - the production of fiber in the
world was 1.25 million fiber miles short of demand last year and is 
expeced to be about the same this year.  IN addition rights of way are
getting rather hard to obtain.  WDM will be a great help but it does
not cause fiber to be run into Seattle.

The cost of putting down fiber is still very high (particularally across
the ponds) and even with WDM the cost of bandwidth will continue
to be high.  Yes the cost should drop (assuming that it is not 
kept high by other things like telco or government policies) but
we ain't going to see "bandwidth too cheap to measure" (as someone
put it on comm-priv 2 years ago) in any timeframe that will let the
Internet community avoid looking at QoS as a very real issue.

Scott
.





More information about the NANOG mailing list