Skip to content

Domain Architecture – Social / Feed

Last Updated: December 29, 2025

1. Purpose

Model post creation, comments, engagement, and feed dissemination with SSE for near real-time updates.

Module Location: bv-social-pulse/

Related Documentation:

2. Services

Service Responsibility Store
post-service CRUD posts Postgres (early) → Mongo (optional)
comment-service CRUD comments Postgres/Mongo
feed-service Materialize feed & SSE push Redis + Kafka
integration-service Outbound webhooks for syndicated content Postgres

3. Protocol Mapping

Use Case Protocol
Post CRUD REST
Feeds aggregated query GraphQL extension
Feed streaming SSE
Post events Kafka
Outbound syndication Webhook
Search (later) OpenSearch

4. Events

Event Purpose
social.post.created.v1 Add to feed pipeline
social.comment.created.v1 Update engagement metrics
social.feed.activity.v1 Aggregated engagement fan-out
social.post.promoted.v1 Boost visibility (optional)

Payload example: 

{
  "eventType":"social.post.created.v1",
  "payload":{
     "postId":"P9",
     "authorId":"U2",
     "tags":["sale","electronics"],
     "createdAt":"..."
  }
}

5. Feed Materialization Options

Strategy: Hybrid push/pull - SSE stream for “latest posts” channel - User-personalized feed (future) built via Kafka Streams state store keyed by userId - Redis lists: feed:global (rolling N), feed:user:{id}

6. Caching

  • Post read: post:{postId}
  • Global feed: list operations
  • Engagement counters: hash feed:engagement:{postId} (increment on comment/like)

7. Security

  • Auth required for posting/commenting.
  • Content moderation pipeline optional; events flagged for ML/AI ingestion.
  • Rate limiting (gateway): posts/min per user.

8. Observability

Metrics:

  • posts_created_total
  • feed_sse_active_connections
  • feed_push_latency_ms
  • engagement_counter_update_failures

Tracing:

  • REST create post -> Kafka publish -> SSE push chain

9. Testing

Layer Focus
Unit Post validation (length, tags)
Integration Event -> feed list update
SSE functional Client receives new post within latency SLO
Performance SSE connection scaling test
Contract GraphQL schema diff gating

10. Implementation Sequence

  1. Post CRUD (REST + Postgres)
  2. Event emission (post.created)
  3. Feed global list consumer + SSE endpoint
  4. Comment service + engagement counters
  5. GraphQL aggregator fields
  6. Webhook integration (syndicated posts)
  7. Personalized feed prototype (user-specific)
  8. OpenSearch indexing (optional later)

11. Interoperability Checklist

  • [ ] Post events schematized & validated
  • [ ] SSE channels documented (naming & reconnect strategy)
  • [ ] GraphQL additions namespaced (no collisions)
  • [ ] No direct dependency on e-commerce code
  • [ ] Webhook format documented for consumers

12. Cost Controls

  • SSE only (skip WebSocket clustering overhead for this domain).
  • Delay search infra until event pipeline stable.

13. Exit Criteria

  • Post creation → SSE push latency < 500ms
  • Feed endpoint returns aggregated data from Redis
  • Comment events trigger engagement counter updates
  • GraphQL federated queries work across Post + User contexts

Architecture References

Implementation Guides

ADRs

Current Status: Active Development ✅
Last Review: December 29, 2025 - Post creation visible via SSE < 500ms median - Engagement counters accurate under concurrent updates - GraphQL consolidation query returning post + engagement fields