roadmap/content/vac/updates/2024-02-05.md

7.8 KiB

title tags date lastmod draft
2024-02-05 Vac weekly
vac-updates
2024-02-05 2024-02-05 false

Vac 2024/02/05

vac:p2p:

  • nimlibp2p:vac:webrtc-transport
    • Fix a bug in Datachannel.read (reading the last message received instead of the first one)
    • Fix a bug due to an Sctp delay (set it to 0ms was the solution)
    • Find a bug in the conception of WebRTCStream. ReadOnce should be Length-prefixed.
      • try to fix it by re-writing ReadOnce, but due to the nature of this proc (inheritance issue) it doesn't work
      • write a RawWebRTCStream to make the length readable without issue
    • Fix a bug with the endianness of the datachannel protocol id
    • E2E Done!
  • nimlibp2p:vac:gossipsub-stagger-send
  • nimlibp2p:vac:maintenance
    • Investigate dependencies issues
      • Found possible problem/s
        • Lack of versioning
        • No major version clamping
        • Using #head
      • Temporary workaround: Clamp/Pin (to git hash) libp2p dependencies' versions
    • Improve documentation [In Progress]
      • Building go-libp2p-daemon
      • Getting Started
      • PR
    • Merge timeout increase
    • Improved checkExpiring
      • Now it'll outpout an error message when it fails due to timeout
        • Not the most visible message

vac:tke:

  • admin/misc:
    • Matty Handoff document finished and share with team on Wed (@Matty)
    • Team Lead Evaluation Criteria finished and share with team on Wed (@Matty)
    • Strengths and development areas for Frederico and Martin, shared with Corey, Daniel, and Jarrad (@Matty)
  • codex:economic-analysis
    • finalize all Codex notion including Dragan's comments to litepaper (@Matty)
    • Wednesday call with Codex team get in sync on next steps
  • status:SNT-staking
    • staking contract implementation becoming a priority, refresh latest progress with SC team (@Martin)
  • nomos:economic-analysis
    • Reading whitepaper and updating TDC (@Frederico)
    • Porting wealth concentration simulation code to GPU to decrease runtimes (@Frederico)
  • waku:economic-analysis
    • Continue Waku Network of Networks design discussion with Franck, concerns around forking (@Martin)
    • Research similar abstract p2p validation protocols (e.g. former Keep Network)

vac:dst:

  • eng-10ktool:vac:bandwidth-test
    • Try to get a stable nim-libp2p version for simulations. Investigated with Alex about building issues with nimble.
    • Analized libp2p metrics, everything normal so far
    • call with p2p team
    • Scale testing for 10K project
      • setup go-waku experiment at scale
      • Successfully simulated a 2,150 node simulation and gathered some basic metrics
      • Modified Kubernetes to allow for more pods to allow for (in theory) scaling to 10k+
      • Failed simulations at 10000 and 5000 nodes - current limits seem to be around ~4800 or so
      • Prometheus is a definite bottleneck - need to switch to a scaled/sharded Prometheus/Thanos setup
      • Attempting one last simulation over the weekend at 4200 nodes
    • Diagnosing 10K project bottlenecks
      • Identified a major potential bottleneck in the form of control plane traffic going over Wireguard / large packet load over WG causing swarm collapse
      • will test the new theory later by re-deploying on Vac Kubernetes with a local control plane + local traffic (while still complying with infra team requirements)

vac:qa:

  • software-testing:waku:test-plans
  • software-testing:waku:interop-testing
    • Relayed messages reach recently started peer with a big delay(@Florin)
    • RLN registration support and tests(@Roman)
  • software-testing:waku:test-automation-go-waku
  • software-testing:waku:test-automation-nwaku
    • Clean and work with Gabriel to verify fix(@Alex)
    • Review lighpush fixes and adjust unit tests(@Alex)
      • Learned how to generate coverage report for NWaku and prepared small PR to have a shortcut(@Roman)

vac:acz:

vac:sc::

vac:nescience:

  • state-separation:vac:state-separation-doc
    • Completed research on SE and DE focusing on security issues while combining both models (Moudy)
    • Rewrote a full version of state update proposal for security and privacy threats (Moudy)
    • Researched address hiding and signature verification and wrote a proposal for address hiding and signature verification (Marvin)
    • Added a report about the security issue and a possible solution(salt mechanism) and investigated about the security of the SE/DE (Ugur)
  • proofsystems:vac:research-existing-proof-systems
  • proofsystems:vac:benchmarks
    • Finished working on arecibo benchmark (Rostyslav)

vac:dr:

  • gsub-scaling:vac:gossipsub-improvements-paper
    • Completed message staggering in the form of weighted message queues
      • Its showing 10% better result than priority queuing, but Async Queue overhead still requires some work
  • zk:codex:storage-proofs-open-problems-review

vac:rfc: