[afnog] QoS/TE

[Bill Woodcock]

      On Wed, 18 Jan 2006, Jim Forster wrote:
    >>> Is QoS/TE in the core really necessary.
    >> Considering the capacity of most modern core equipment,
    >> I tend to say avoid QoS there.  It's usually cheaper
    >> to upgrade the core than to deploy QoS.
    > I would tend to agree, although most of the thought and analysis has  
    > been done for places with fiber based cores.   In these places the  
    > cost of additional bandwidth for core networks is generally cheaper  
    > than additional bandwidth for access networks.

Absolutely agreed...  QoS imposes its own huge set of failure modes and 
poor performance and customer annoyance, beyond just the cost and risk of 
it.  And it never _solves_ a problem, it just _mitigates it in a minority 
of cases_.  You're essentially always better off putting the same amount 
or less effort into actually _solving_ the problem.  I know it may not 
seem that way, but you're very lucky in Africa, in that the ratio of cost 
of capital to cost of installing fiber (yes, including "special services" 
and political difficulties) is far lower than in, say, the U.S. or Europe.  
That means that, viewed overall, you've got even more reason to solve your 
problems by digging a ditch and throwin in fiber than equivalent operators 
elsewhere.  Take full advantage of that!

                                -Bill