Matthieu Baerts

Hello, Thank you for reporting this. It would be interesting to analyse what's the limitation for both TCP and MPTCP: CPU (any visible bottlenecks?), buffers (MPTCP might require more space),...

@roman-shuhov thank you for the experimentations! > [@matttbe](https://github.com/matttbe) do you think this are reasonable perf numbers for 1 connection comparison between TCP and MPTCP? I would expect them to be...

@roman-shuhov Thank you for the experiment. The output from `ss -itm` is very "light". Did you strip some info? Or do you use an old version? (I remember the v6.7...

> > GRO/GSO > > There is gro/gso settings on the NIC with 64KB max_size. I bumped both of them to 512KB (hard limit) and it improved throughput proportionally for...

@roman-shuhov & @cpaasch-oai : thank you for the extra tests, and for having explored hw-GRO. So the difference might be due to what is described in the [FAQ](https://www.mptcp.dev/faq.html#is-gro-supported-with-mptcp)? This part...

Thank you for the new results! > * for MPTCP with 1 subflow, why there is an excessive amount of Retransmits. It's also visible without "fixed" window. I didn't check...

Note that we briefly discussed this issue during the last [Meeting](https://github.com/multipath-tcp/mptcp_net-next/wiki/Meetings). Here are some notes: - HW GRO seems to be the culprit: it would be good to start discussing...

Hi @suyuan168, Thank you for the (long) bug report, and sorry for the delay. If I understood correctly (if you use AI, please ask it to use more concise descriptions):...

> Later, if one 5G path flaps (or only one 5G remains for a while), the remaining active 5G capacity temporarily cannot sustain the 10 Mbps app rate and the...

Hi @anthonyryan1 Sorry for the delay, and for having missed the question on IRC. `signal` + `fullmesh` flags don't make sense, but indeed, I guess it can when the `subflow`...