search

panrg / path-properties

A Vocabulary of Path Properties
Other
1 stars 3 forks source link

Wireless properties, MTU #4

Closed renghardt closed 5 years ago

renghardt commented 5 years ago

This (really short) Pull Request substitutes the Wireless properties with the more abstract data rates based on RFC 8175. Also, I substituted "available bandwidth" (capacity) with "maximum data rate" in the process, because it's actually the same thing, just available bandwidth (capacty) is supposed to apply to the entire path instead of just one link. Now this is the theoretical maximum of all links on the path, which might be different from the actually achievable throughput on this path. But I'm not sure we will be able to specify the actually achievable throughput on a path with the definition of path on a per-packet basis and with flows... maybe the flow can have a maximum throughput later? But that's future work.

In this PR, I'm not sure if we need to reference RFC 8175? It inspires our definition, but then again, it's not like they provide an entirely novel thing... And DLEP would be a way to gather these properties, but how to gather properties is out of scope for our draft.

Any opinions?

cyrill-k commented 5 years ago

I also like maximum data rate more than "achievable throughput" as the achievable throughput depends on e.g., the protocol and congestion control algorithms used. I think I would cite RFC8175 to show that our wireless path properties are also defined elsewhere in this way.

renghardt commented 5 years ago

Okay, I added a sentence referencing RFC 8175. Also, I substituted "end-to-end bandwidth" in the text, because we don't have that anymore, with "end-to-end latency", because we will most likely keep this one. Does this work?