Event & Message Fabric

Responsibilities

• Provide async and real-time messaging with delivery guarantees.
• Enable pub/sub, streaming, and queues with backpressure.
• Standardize event schemas and topic governance.

Key interfaces

• Topic naming and schema registry (where used).
• Consumer group semantics and replay policies.
• Dead-lettering and retry strategies.

Operational signals

These are the measurements that tell you whether this layer is healthy in production.

• Lag/backlog by topic/consumer group.
• Delivery failure rate and redelivery storms.
• Throughput and saturation of brokers/workers.

Failure modes

• Poison messages causing stuck consumers.
• Unbounded retries producing traffic amplification.
• Schema changes breaking consumers.

Production readiness checklist

• Use DLQ + quarantine patterns; cap retries.
• Version event schemas and validate compatibility.
• Monitor lag and autoscale consumers safely.