search

ethereum / eth2.0-pm

ETH2.0 project management
Other
260 stars 75 forks source link

Eth2 Networking Call 3 #131

Closed djrtwo closed 4 years ago

djrtwo commented 4 years ago

Call 2

Meeting Date/Time:

Meeting Date/Time: Wednesday 2020/3/4 at 13:00 GMT

Meeting Duration:

1.25 hours

Agenda

  1. Opening
  2. Updates
  3. Spec items
  4. Discussion of items that are under-{researched, tested, resourced}
  5. Open discussion
  6. Closing remarks and plans for next call or other organizational items
ralexstokes commented 4 years ago

s\Wednesday 2020/1/29\Wednesday 2020/3/4\

AgeManning commented 4 years ago

Item to discuss:

The process of finding and connecting peers goes like:

A connecting node will know the ENR of the peer it's connecting to. In some cases, if a peer has found the ENR via discv5, the receiving peer will have had some communication with the connecting peer and may/may not have its ENR in it's local node routing table.

In other cases, a peer may get ENR's from boot-node lists or websites etc and connect to a listening peer on tcp via libp2p. In this case the receiving peer does not know the connecting peer's ENR. This wasn't an issue until we needed the ENR for further logic. Specifically the subnet Bitfield. We would like to know ENR's of all connected peers to know if they are suitable for specific subnet subscriptions.

I'm suggesting we add the ENR in the status message, or add an RPC method that gives us the ENR, or to take it one step further, in the protocol negotiation of RPC have a connecting peer send it's ENR. Having the ENR will also help discovery as nodes can inject the ENR into their local node table for connecting peers.

benjaminion commented 4 years ago

My notes from the call.

nkeywal commented 4 years ago

Here are the notes of our post call discussion: Participants: Dmitriy, Mikerah, Cayman, Protolambda, Nicolas, Marin

We put down a possible plan, and wrote a check list

Proposed plan

What we have today: testnet ready network

Put in place an incentivized network

Ethereum Phase 0: genesis ready network Ethereum Phase 1: analysis of the performances in real deployment, choice of requirements; attacks Ethereum Phase 2: network next gen

Checklist

Category Point Importance Comment
Attacks DoS Phase 0: P1 Target: Cryptoeconomically impossible
Attacks Eclipse Phase 0: P1 Target: Cryptoeconomically impossible
       
Scalability Attestations: aggregation with 1K nodes Phase 0: P1  
  Attestations: aggregation with 20K nodes Phase 2: P1  
       
Privacy See if it’s possible to deanomyze the network Phase 0: P1