[e2e] QoS vs Bandwidth Overprovisioning

[demir]

> Unfortunately, when you start to check on e2e delay, it can be quite
> large. This's because end-user packets go over multiple carriers and
> multiple inter-domain connection. There is no guarantee on end-to-end
> delay in the Internet today. Period!

There is an end-to-end delay guarantee in the Internet today, and it is
the worst-case delay from one end to other. However, Internet QoS
architecture provides only "best-effort" service. In your example, it is
"best-effort" delay.  Applications (customers) who are using this
service are satisfied with this service. We know that there are some
applications are not satisfied with thise service. As a result, if one who
wants to support them should use more sophisticated QoS
architecture. There are factors, i.e., scalability, deployment, security,
economical,  that prevented us from using/deploying these QoS
architectures. The problem in designing a QoS architecture is how we could
optimize network resources for demanding traffic and what the desired QoS
is (how we define the QoS; resulting with related services). To me,
"best-effort" Internet, "Intserv" and "diffserv" architectures, all, are
emerged with all these factors in mind; leading different/various 
mechanisms to be integrated to create different/various QoS architectures.

Alper K. Demir