Skip to content

P2.md

Latest commit

 

History

History
131 lines (94 loc) · 4.3 KB

P2.md

File metadata and controls

131 lines (94 loc) · 4.3 KB
Raw

P2 — Experimental L1 stream composition

Paired Chinese version: P2.zh-CN.md

Revision: 2026-08-22

Public guides: Run an L1 node · Linux runtime

1. Status and boundary

P2 records an experimental composition in which selected geth or Prysm TCP streams are carried by the same persistent web3:// transport used by applications.

This is laboratory work, not the public L1 joining contract. Public operators still join CoNET L1 through the documented geth and Prysm P2P path. P2 does not claim that existing validators, beacon nodes, or execution clients have moved to Layer Minus.

2. Reused application transport

The composition reuses the P1 stream contract:

local L1 process
    ⇅ TCP
conet-l0d client endpoint
    ⇅ encrypted Layer Minus session
conet-l0d --proxyDuplex
    ⇅ TCP
remote L1 process

The application identity remains a wallet address or exact BeamioTag:

web3://<wallet-or-tag>:8400
web3://<wallet-or-tag>:4200

The ports are application metadata:

Port Laboratory use
8400 geth devp2p TCP stream
4200 Prysm libp2p TCP stream

No new SI command family or public endpoint is introduced. The session uses the existing Layer Minus entry/mailbox attachment path plus versioned application-session messages.

3. Laboratory record

The August 2026 two-host experiment used:

Role Public host Overlay endpoint
Spoke 74.208.224.45 100.64.0.5
Hub 198.251.77.98 100.64.0.6

The hub published local L1 services with --proxyDuplex. The spoke selected the hub with a web3:// target and exposed a local client endpoint. Existing geth and beacon processes then used ordinary TCP against those local endpoints.

The experiment demonstrated:

  1. an accepted wallet-addressed duplex offer;
  2. continuous encrypted frames in both directions;
  3. a stable TCP connection for the selected L1 service;
  4. reconnect after an application-stream interruption; and
  5. unchanged geth and Prysm peer identities.

The specific addresses above are laboratory facts, not universal defaults.

4. Acceptance evidence

P2 is considered active only when all applicable evidence agrees:

Layer Evidence
Runtime conet-l0d status reports the expected server and client endpoints
Layer Minus logs show accepted duplex sessions and continuing encrypted frames
TCP the selected local and remote service connections remain established
Geth net_peerCount and the expected enode confirm the execution peer
Prysm /eth/v1/node/peer_count and peer identity confirm the consensus peer
Sync head slot rises and sync distance falls over time

A running daemon alone is not proof that the L1 peer is connected.

5. Reconnect and recovery

Recovery proceeds from the application transport outward:

  1. verify the local service and the published upstream are listening;
  2. verify the wallet route and duplex session;
  3. verify encrypted frames continue in both directions;
  4. verify the L1 client has an established TCP peer; and
  5. only then diagnose geth or Prysm peer state.

Restarting geth, beacon-chain, or validator remains an explicit operator decision and requires prior authorization. Transport diagnosis must not silently restart chain infrastructure.

6. Out of scope

P2 does not:

  • replace the public geth/Prysm node-joining guide;
  • define a general IP network;
  • make browser clients operate L1 consensus software;
  • add a new public domain or SI wire command;
  • promise production availability for L1-over-L0; or
  • turn SilentPass into an L1 consensus transport.

Earlier datagram and discovery experiments remain engineering history. They are not part of the public web3:// application contract.

7. Relationship to the product

The product hierarchy is:

  1. Layer Minus (L0) supplies encrypted, wallet-routed transport.
  2. web3:// defines the application locator and session contract.
  3. conet-l0d provides the Linux server/client runtime.
  4. P2 explores one optional Linux-to-Linux use of that runtime for selected L1 TCP streams.

The application protocol remains useful to ordinary services and browser or native clients regardless of the P2 experiment.