DST summary 2025Q3

I will present here some of the work that has been done from DST team in 2025Q3. These are the major things, also there are other smaller bugs and issues opened that are not mentioned here… Everything is public in the DST notion page.

  • Waku:

    • We did a very interesting analysis over Discv5 bandwidth consumption.
    • We finally have a working setup for JS-waku in the lab, ready for more complex tasks!
    • We are creating a Docker image to interact with Waku REST API. We will add new endpoints depending on what we need, but hopefully this will be useful for others.
    • Links:
  • Status:

    • After a lot of work, we faced some issues with status discovery in the chat protocol benchmarks. This exploded into multiple lines of research, where bugs in functional tests and issues in peer exchange mechanisms were discovered. Notably, a discrepancy in peer-exchange implementation agains the specs was found (Thanks to everyone involved in this)
    • Disovery in relay and light nodes is mixed, and we are performing the benchmark again, at the beginning of Q4.
    • Links:
  • Nim-libp2p & ACZ:

    • Reports to keep track of the stability of every multiplexor. Quic is still presenting some issues that are being investigated by the P2P team. Other than that, this also helped us to discover some instabilities in the lab, that are already fixed.
    • In Mix protocol, we revisited the previous analysis and did a new one with up to date changes. We confirmed the previous bug was fixed, where messages were being relayed through Gossipsub and mix at the same time, where this should not happen. Also, we discovered other issues related with Mix message propagations and message reliability.
    • We are preparing an interop scenario in Kuberentes and Shadow for testing several implementations of libp2p.
    • Links:
  • Nomos:

2 Likes