Interesting idea, AF_XDP as a channel to TCP offload. If I understand this properly, the number of applications using this would be limited to the number of hardware queues the NIC has and a single AF_XDP socket to the application may be a bottleneck.

https://www.sipanda.io/post/episode-iv-a-new-hope-for-tcp-offload-that-is

https://blog.ipspace.net/2024/10/comparing-ip-clnp-addressing/

https://www.sipanda.io/post/segmentation-offload-and-protocols-let-s-be-friends

https://jack.wrenn.fyi/blog/safety-goggles-for-alchemists/

Over on the Preseem blog I posted a bit of fiber data analysis comparing signal strength vs fiber length for Calix, Adtran and Ubiquiti.

We have a lot of data, just need more time to explore it.

https://preseem.com/2024/09/ont-receive-power-fiber-length

BGP, EVPN, VXLAN, or SRv6?

This is roughly what we’ve been doing with some internal Rust work as well.

https://blog.sulami.xyz/posts/sync-core-async-shell

https://quic.video/blog/never-use-datagrams

https://quic.video/blog/quic-powers

https://phys.org/news/2024-05-carbon-pricing-major-meta.html