Understanding "Client ID: UNKNOWN, IP disconnected ungracefully" in HiveMQ

🔍 What Does This Log Mean?

You might see logs like:

This means:

• A TCP connection was accepted by HiveMQ.

• No MQTT CONNECT packet was ever received.

• The connection was closed—likely by the client or intermediary.

• Therefore, HiveMQ doesn't know the client ID (UNKNOWN) and flags it as ungraceful disconnect.

💡 Common Causes

1. TCP Load Balancer Health Checks

In AKS or other cloud setups, these logs are often caused by:

• Load balancers performing TCP-level health checks that open and immediately close connections.

• These do not complete an MQTT handshake.

2. Network Interruptions or Timeouts

Especially relevant in Kubernetes clusters:

• Pods restarting.

• Network plugin issues.

• Intermittent node-to-node communication dropouts.

3. TLS Termination or Misconfiguration

• If TLS is terminated at a proxy/load balancer and not forwarded correctly, HiveMQ may see incomplete handshakes.

🔧 How to Diagnose

• Check your load balancer settings. For Azure:

  • Use HTTP(S) probes instead of TCP, if possible.

  • Exclude MQTT listener ports from health checks.

• Verify no security groups/firewalls are dropping packets.

• Enable DEBUG logs temporarily for the event_log or network layers to confirm patterns.

🧼 Should You Worry?

No—these logs are informational and can be safely ignored if caused by expected probes.

 Related articles