feat(us-017): P2P gossip, NAT-traversal relay node, and TLS

- packages/p2p: identity (peer_id from sha256 of RSA pubkey), TLS cert
  generation with SHA-256 fingerprint, GossipClient (WSS PubSub with
  per-topic handlers, dedup by msg_id, auto-reconnect), MdnsDiscovery
  (zeroconf optional dependency, graceful no-op fallback)
- packages/relay: new meshnet-relay package — RelayServer (asyncio +
  websockets) with gossip fanout hub, circuit relay proxy for NAT traversal,
  peer registry; meshnet-relay CLI
- packages/p2p/relay_bootstrap.json: team relay bootstrap list
- Tracker: _NodeEntry gains relay_addr, cert_fingerprint, peer_id; both
  register and heartbeat handlers read and store these optional fields
- docs/adr/0010 already written (previous commit)
- conftest.py: packages/relay added to sys.path
- 18 new tests; 115 passed total, 1 skipped (no regressions)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Dobromir Popov
2026-06-29 17:58:16 +03:00
parent 65f3ee6a85
commit a2258d3df4
14 changed files with 1259 additions and 0 deletions

View File

@@ -56,6 +56,7 @@ class _NodeEntry:
"score", "vram_bytes", "ram_bytes", "quantizations",
"benchmark_tokens_per_sec", "quantization", "managed_assignment",
"pending_directives", "last_heartbeat", "tracker_mode",
"relay_addr", "cert_fingerprint", "peer_id",
)
def __init__(
@@ -76,6 +77,9 @@ class _NodeEntry:
quantization: str | None = None,
managed_assignment: bool = False,
tracker_mode: bool = False,
relay_addr: str | None = None,
cert_fingerprint: str | None = None,
peer_id: str | None = None,
) -> None:
self.node_id = node_id
self.endpoint = endpoint
@@ -93,6 +97,9 @@ class _NodeEntry:
self.quantization = quantization
self.managed_assignment = managed_assignment
self.tracker_mode = tracker_mode
self.relay_addr = relay_addr
self.cert_fingerprint = cert_fingerprint
self.peer_id = peer_id
self.pending_directives: list[dict] = []
self.last_heartbeat: float = time.monotonic()
@@ -607,6 +614,9 @@ class _TrackerHandler(http.server.BaseHTTPRequestHandler):
return
tracker_mode = bool(body.get("tracker_mode", False))
relay_addr = body.get("relay_addr") or None
cert_fingerprint = body.get("cert_fingerprint") or None
peer_id = body.get("peer_id") or None
node_id = str(uuid.uuid4())
entry = _NodeEntry(
@@ -626,6 +636,9 @@ class _TrackerHandler(http.server.BaseHTTPRequestHandler):
quantization=quantization,
managed_assignment=not explicit_shard,
tracker_mode=tracker_mode,
relay_addr=relay_addr,
cert_fingerprint=cert_fingerprint,
peer_id=peer_id,
)
with server.lock:
self._purge_expired_nodes()
@@ -643,6 +656,13 @@ class _TrackerHandler(http.server.BaseHTTPRequestHandler):
def _handle_heartbeat(self, node_id: str):
server: _TrackerHTTPServer = self.server # type: ignore[assignment]
body: dict = {}
content_length = int(self.headers.get("Content-Length", 0))
if content_length > 0:
try:
body = json.loads(self.rfile.read(content_length))
except Exception:
pass
with server.lock:
self._purge_expired_nodes()
entry = server.registry.get(node_id)
@@ -650,6 +670,12 @@ class _TrackerHandler(http.server.BaseHTTPRequestHandler):
self._send_json(404, {"error": "node not found"})
return
entry.last_heartbeat = time.monotonic()
if body.get("relay_addr"):
entry.relay_addr = body["relay_addr"]
if body.get("cert_fingerprint"):
entry.cert_fingerprint = body["cert_fingerprint"]
if body.get("peer_id"):
entry.peer_id = body["peer_id"]
_rebalance_model_locked(server, entry.model or "stub-model")
directives = list(entry.pending_directives)
entry.pending_directives.clear()