Files
buzz/mobile/lib/shared/relay/relay_socket.dart
Luke Tornquist 974495fd4a fix(mobile): resume reconnect replay from the last live point and spread reconnects (#8659)
## Summary

A relay rolling deploy closes every socket with 1012. On reconnect,
mobile replayed each live subscription's **original filter** (startup
`since`, full `limit`) unless that subscription had received an event,
so quiet subscriptions re-read history back to app start every time.
Reconnects used a fixed 1s delay with no jitter. The delay also reset to
1s on every successful connect, so a client bounced from pod to pod
during a deploy kept coming back at the same moment as every other
client.

**This PR makes two changes:**

1. **Resume cursor.** If a subscription reached EOSE on a connection, it
records `liveUntil`: the time of the last frame received from the relay
on that connection, capped at the clock at close. The replay then uses
`since = max(original since, lastSeen − 5s, liveUntil − 960s)`, and
never goes past the filter's `until`.
- **Why 960s:** the relay accepts `created_at` up to ±900s from its
clock (`MAX_TIMESTAMP_DRIFT_SECS` in
`crates/buzz-relay/src/handlers/ingest.rs`). An event stored after the
close can therefore be backdated by that much, and the extra 60s covers
client clock skew.
- **Only relay frames count as liveness.** Close and error paths never
extend `liveUntil`, so a close noticed after the device sleeps doesn't
overstate how long the subscription was live.
- **Ambiguous EOSE is ignored.** If a subscription has a second REQ on
the same socket, that subscription doesn't advance `liveUntil` for the
rest of the socket. A CLOSED ends liveness for the rest of that socket,
because the retry is a second REQ. Subscription IDs are never reused, so
an old EOSE can't mark a re-subscribe as live.
- **No cursor:** a subscription that never reached EOSE keeps its
previous cursor, or replays as before.
2. **Reconnect spreading.**
- After a 1012 or 1001 close, the first attempt waits a uniform random
delay in [0, 10s]. A client already backing off waits the longer of that
delay and its backoff.
   - Ordinary backoff gets symmetric ±25% jitter, capped at 30s.
- The backoff resets only after a connection stays up for 60s, not on
connect. This matches desktop.
- Foreground resume and a manual `reconnect()` still reset to the base
delay, as before.

**Simpler?** The pure reconnect math (cursor, jitter, spread, stability)
lives in a new `relay_reconnect_policy.dart` and is table-tested on its
own. `relay_session.dart` was near its 1200-line cap, so its five small
private bookkeeping classes moved verbatim to
`relay_session_subscriptions.dart`. Net concept count is unchanged.

**Identified debt (not changed here):**
- Dart's WebSocket ping/pong isn't visible to the session, so a
long-idle but healthy connection that drops replays back to its last
application frame − 960s. This is still bounded, and much cheaper than
replaying from app start.
- The unread catch-up in `channels_provider.dart` runs one REQ per
active channel (`since = readAt + 1`, `limit: 1000`) on reconnect and
every 60s.
- A retryable CLOSED is retried with a 30s cap, no jitter and no attempt
limit.
- `lastSeenCreatedAt` advances on stored events received before EOSE. If
a relay returns stored events newest-first and the socket dies before
EOSE, the next replay can skip older stored events.
- The `lastSeen − 5s` term keeps its existing gap for events backdated
after a close.

### Related issue
N/A. This is part of a set of reconnect-load reductions. Companion PRs:
relay read-pool warm (#8651), desktop (#8652), agent harness (#8653).

### Testing
- Mobile gate on a Linux workstation, at the head SHA:
  - `dart format --set-exit-if-changed`: pass.
  - `flutter analyze`: no issues.
  - Gateway recipe tests: pass.
- `flutter test`: 3881 pass, 1 fail.
`push_gateway_origin_validator_test` fails the same way on `main` on
that host, due to a host git-config error.
- New `relay_reconnect_resume_test.dart`:
  - pure table tests for the cursor and the delay
  - a real-socket test that reads the 1012 close code
- journeys through `RelaySessionNotifier`. They check that a close or
error after an hour asleep doesn't extend liveness, that a second REQ or
a delayed EOSE for an unsubscribed ID can't advance the cursor, and that
a backdated event inside the 960s window is still replayed.
- 22 production mutations, each failing the new tests. They cover every
cursor term, the 960s window, the `until` clamp, the liveness cap, the
ambiguous-EOSE guard, CLOSED handling, jitter, the 1012 spread and
floor, the backoff reset, and close-code capture.
- `fake_relay.dart` now draws a fixed jitter midpoint, so existing
journey tests stay deterministic.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Signed-off-by: Luke Tornquist <tornquist@squareup.com>
Co-authored-by: Elrond <28d6302a099e5225b02c4155ac4236e4912603df2ab08dbfc2f4fef08ce598c8@buzz.block.builderlab.xyz>
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-28 17:35:12 -07:00

284 lines
7.7 KiB
Dart

import 'dart:io';
import 'dart:async';
import 'dart:convert';
import 'package:flutter/foundation.dart';
import 'package:nostr/nostr.dart' as nostr;
import 'package:web_socket_channel/io.dart';
import 'package:web_socket_channel/web_socket_channel.dart';
import 'nostr_models.dart';
import 'relay_websocket.dart';
import 'package:buzz/shared/crypto/event_signing.dart';
/// Low-level websocket connection with NIP-42 authentication.
///
/// Handles the raw websocket lifecycle: connect, authenticate via NIP-42
/// challenge/response, send/receive JSON frames, and disconnect.
///
/// Does NOT handle reconnection — that is [RelaySessionNotifier]'s job.
enum SocketState { disconnected, connecting, authenticating, connected }
class RelayAuthRejectedException implements Exception {
final String message;
const RelayAuthRejectedException(this.message);
@override
String toString() => 'Relay authentication rejected: $message';
}
Exception classifyRelayAuthFailure(String message) {
if (message.startsWith('error:')) return Exception(message);
return RelayAuthRejectedException(message);
}
class RelaySocket {
/// Interval for sending a ping and awaiting its pong before disconnecting.
static const pingInterval = Duration(seconds: 30);
@visibleForTesting
static Duration debugPingInterval = pingInterval;
final String _wsUrl;
final String? _nsec;
final void Function(List<dynamic> message) _onMessage;
final void Function() _onConnected;
final void Function(Object? error) _onDisconnected;
WebSocketChannel? _channel;
StreamSubscription<dynamic>? _subscription;
SocketState _state = SocketState.disconnected;
Completer<void>? _authCompleter;
Timer? _authTimeout;
String? _pendingAuthEventId;
int? _closeCode;
SocketState get state => _state;
/// The WebSocket close code the relay sent, once it has closed the
/// connection; null if the connection failed without one.
int? get closeCode => _closeCode;
/// Scoped assertion supplied by the connection owner before connecting.
Map<String, String> headers = const {};
RelaySocket({
required String wsUrl,
required String? nsec,
required void Function(List<dynamic> message) onMessage,
required void Function() onConnected,
required void Function(Object? error) onDisconnected,
}) : _wsUrl = wsUrl,
_nsec = nsec,
_onMessage = onMessage,
_onConnected = onConnected,
_onDisconnected = onDisconnected;
/// Connect to the relay and complete NIP-42 authentication.
Future<void> connect() async {
if (_state != SocketState.disconnected) return;
_state = SocketState.connecting;
try {
_channel = IOWebSocketChannel(
openRelayWebSocket(
Uri.parse(_wsUrl),
headers: headers,
pingInterval: debugPingInterval,
),
);
await _channel!.ready;
} catch (e) {
_state = SocketState.disconnected;
_onDisconnected(e);
return;
}
// The channel may have been disposed while we were awaiting ready
// (e.g. provider rebuild triggered dispose() concurrently).
if (_channel == null) {
_state = SocketState.disconnected;
return;
}
_state = SocketState.authenticating;
_authCompleter = Completer<void>();
_subscription = _channel!.stream.listen(
_handleRawMessage,
onError: (Object error) {
_failAuth(error);
_resetConnection();
_onDisconnected(error);
},
onDone: () {
_closeCode = _channel?.closeCode;
_failAuth(null);
_resetConnection();
_onDisconnected(null);
},
);
// Wait for auth to complete (or timeout).
_authTimeout = Timer(const Duration(seconds: 8), () {
if (_authCompleter != null && !_authCompleter!.isCompleted) {
_authCompleter!.completeError(
TimeoutException('NIP-42 auth timed out after 8s'),
);
}
});
try {
await _authCompleter!.future;
_authTimeout?.cancel();
_state = SocketState.connected;
_onConnected();
} catch (e) {
_authTimeout?.cancel();
await disconnect();
_onDisconnected(e);
}
}
/// Send a raw JSON array over the websocket.
void send(List<dynamic> payload) {
_channel?.sink.add(jsonEncode(payload));
}
/// Gracefully close the connection.
Future<void> disconnect() async {
_resetConnection();
final channel = _channel;
_channel = null;
if (channel != null) {
await channel.sink.close();
}
}
void dispose() {
_resetConnection();
_channel?.sink.close();
_channel = null;
}
void _resetConnection() {
_state = SocketState.disconnected;
_subscription?.cancel();
_subscription = null;
_authTimeout?.cancel();
_authTimeout = null;
_pendingAuthEventId = null;
}
void _failAuth(Object? error) {
if (_authCompleter != null && !_authCompleter!.isCompleted) {
_authCompleter!.completeError(error ?? Exception('Connection closed'));
}
}
void _handleRawMessage(dynamic raw) {
final String text;
if (raw is String) {
text = raw;
} else {
return; // Binary frames are not part of the Nostr protocol.
}
final List<dynamic> data;
try {
data = jsonDecode(text) as List<dynamic>;
} catch (_) {
return; // Malformed JSON.
}
if (data.isEmpty) return;
final type = data[0] as String;
switch (type) {
case 'AUTH':
_handleAuthChallenge(data);
case 'OK':
_handleOk(data);
default:
// Pass EVENT, EOSE, NOTICE, etc. upstream.
_onMessage(data);
}
}
/// Handle the relay's AUTH challenge: sign a kind:22242 event and respond.
void _handleAuthChallenge(List<dynamic> data) {
if (data.length < 2) return;
final challenge = data[1] as String;
if (_nsec == null) {
_failAuth(Exception('No nsec available for NIP-42 auth'));
return;
}
try {
// Decode bech32 nsec to hex private key.
final privkeyHex = nostr.Nip19.decode(payload: _nsec).data;
if (privkeyHex.isEmpty) {
_failAuth(Exception('Invalid nsec'));
return;
}
// Build the auth tags.
final tags = <List<String>>[
['relay', _wsUrl],
['challenge', challenge],
];
// Create and sign the kind:22242 AUTH event.
final event = signNostrEvent(
kind: EventKind.auth,
content: '',
tags: tags,
secretKey: privkeyHex,
);
_pendingAuthEventId = event.id;
send(['AUTH', event.toMap()]);
} catch (e) {
_failAuth(e);
}
}
@visibleForTesting
void debugHandleOkForTest(List<dynamic> data) => _onMessage(data);
/// Handle OK frames. During auth, complete the auth flow.
void _handleOk(List<dynamic> data) {
if (data.length < 3) return;
final eventId = data[1] as String;
final accepted = data[2] as bool;
// Check if this OK is for our pending AUTH event.
if (_pendingAuthEventId != null && eventId == _pendingAuthEventId) {
_pendingAuthEventId = null;
if (accepted) {
if (_authCompleter != null && !_authCompleter!.isCompleted) {
_authCompleter!.complete();
}
} else {
final message = data.length > 3
? data[3] as String
: 'Auth rejected by relay';
_failAuth(classifyRelayAuthFailure(message));
}
return;
}
// Pass non-auth OK frames upstream for pending event tracking.
_onMessage(data);
}
}
/// HTTP upgrade failures retain their status through WebSocketChannel.
bool relayUpgradeRefused(Object? error) {
if (error is WebSocketChannelException) error = error.inner;
return error is WebSocketException &&
(error.httpStatusCode == 401 || error.httpStatusCode == 403);
}