On Wed, Aug 14, 2024 at 3:32 PM Dominik Dobrowolski <[email protected]> wrote: > In your article you mention two costs: > > 1. Cost of transport > 2. Cost of hardware > > Both are included into the transit price we pay. If it was unprofitable, > transit prices would be higher to cover the costs.
Howdy, Doesn't work that way. Look up "tragedy of the commons," and if it's not obvious ask me to explain further and I will. > Keep in mind that a full table in FIB is not necessary for DFZ access, there > are plenty of people > running software routers capable of high throughput (especially with VPP, > DPDK) It's not about what you spend or might be able to spend. It's about what everybody else spends. I wrote some of the software AWS runs using DPDK. It's a great platform. It's a great tool for building custom solutions at the network edge. Their network exterior participating in the DFZ runs on Cisco and Juniper big-iron like everybody else and your announcement from wherever you are in the world consumes a slot in their routing tables. Unless it happens to be with their BYOIP program, neither you nor anybody else pays them for that consumption. > Holding the growth of internet just to keep the life of routers longer is > 2008 mentality, just like the article mentioned. The exact input numbers have changed since the paper was written in 2008. The math equations have not. Regards, Bill Herrin -- William Herrin [email protected] https://bill.herrin.us/ _______________________________________________ ARIN-PPML You are receiving this message because you are subscribed to the ARIN Public Policy Mailing List ([email protected]). Unsubscribe or manage your mailing list subscription at: https://lists.arin.net/mailman/listinfo/arin-ppml Please contact [email protected] if you experience any issues.
