#221 Presence recovery after gateway restart FIXED

[felladaniel36-hash]

Description

This PR resolves an issue where online presence tracking and room subscriptions were lost during a gateway (backend instance) restart.

Specifically, it introduces three key changes:

1. Gateway Restart Protection: Hooks server shutdown lifecycle events (SIGTERM, SIGINT, httpServer.close(), io.close()) to set an isShuttingDown flag. During a server shutdown or namespace restart, socket.on('disconnect') skips asserting setOffline(), preserving active user presence mappings in Redis.
2. Boot Room Subscription Reconciliation (reconcileBoot): Adds a reconciliation routine that runs on gateway boot once the Socket.IO Redis adapter attaches. It scans active presence:* keys in Redis, queries PostgreSQL for user conversation memberships, and rebuilds room subscriptions via io.in(socketId).socketsJoin(conversationId).
3. Stale/Duplicate Socket Cleanup (cleanupStaleSockets): On socket connection, heartbeat, and disconnect, existing socket IDs stored in presence:{userId} are checked against active sockets in the cluster (io.in(sid).fetchSockets()). Dead sockets from prior restarts are purged immediately, preventing duplicate presence entries.

Type of change

• Bug fix

• New feature

• Documentation update

• Other

Checklist

• I have read the contributing guidelines

• I have tested my changes locally

• My code follows the project's coding standards

CLOSE #221

[drips-wave]

@felladaniel36-hash Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits

[codebestia]

Hello @felladaniel36-hash
Please fix the conflicts and ensure the CI passes.
Thanks