Message Queue Health Check Component Design

Message Queue Health Check

Queue health checks verify that your asynchronous messaging infrastructure is functioning. Queue issues can cause silent data loss or processing delays.

Response Component

{
  "queue": {
    "status": "UP",
    "duration": "8ms",
    "message": "RabbitMQ connected, queues operational",
    "details": {
      "type": "rabbitmq",
      "connection": "open",
      "channels": 3,
      "queues": {
        "orders": { "messages": 42, "consumers": 2 },
        "notifications": { "messages": 0, "consumers": 1 }
      }
    }
  }
}

What to Check

1. Connection status: Is the broker reachable?
2. Queue depth: Are messages piling up?
3. Consumer count: Are consumers attached and processing?
4. Dead letter queue: Are messages failing processing?
5. Consumer lag (Kafka): How far behind are consumers?

Health Thresholds

Metric	Healthy	Degraded	Unhealthy
Connection	Open	Reconnecting	Failed
Queue depth	< 1000	1000-10000	> 10000
Consumer count	> 0	0 (briefly)	0 (extended)
Consumer lag	< 100	100-10000	> 10000
DLQ messages	0	< 10	> 10

Producer vs Consumer Health

Producer services should check:

• Connection to broker
• Ability to publish messages
• Queue existence

Consumer services should check:

• Connection to broker
• Consumer registration
• Processing throughput
• Dead letter queue size

Kafka-Specific Checks

For Kafka, include consumer group lag per partition and check that the consumer group is active. High lag indicates consumers are falling behind, which may mean the service cannot process real-time data effectively.