25 Jan
2021
25 Jan
'21
5:49 a.m.
On 1/24/21, Jon Steinhart <jon(a)fourwinds.com> wrote:
So I never liked Apollos much. What I was referring
to was Apollo's claim
that their token-ring network performed better for large numbers of nodes.
And they were correct. However, they didn't consider the eventually
invention of switches that solved the problem.
A problem that shouldn't have ever been there in the first place. When
I was at EDS, we did a lot of benchmarks against token-ring vs.
CSMA-CD. Token-ring was slower than CSMA-CD until the traffic got to
be more than about 10% of capacity - then the collision detection
exponential backoff algorithm would clobber the network. The argument
that "well, we will never get above that anyway, so we want the
fastest we can get" sort of short-sightedness won the day. It wasn't
until switches and virtual LANs came into existence that (as you said)
solved the problem.