Rule Engine Monitoring

Monitor Rule Engine health through the built-in Statistics dashboard and investigate individual node failures using Debug mode. These are the two primary tools for troubleshooting message processing issues in ThingsBoard.

The Rule Engine statistics dashboard tracks message throughput and surfaces processing failures and exceptions across all queues and rule engine partitions.

To open the dashboard:

Go to API Usage in the top-right menu → click Rule Engine executions → click View statistics in the top-right corner of the chart.

The dashboard contains three panels:

Realtime chart. Shows per-partition message counts broken down by outcome:

Series	What it counts
Successful	Messages that completed processing without error
Permanent Failures	Messages that exhausted all retries and were discarded
Processing Failures	Messages that failed during rule chain execution

Each queue partition is tracked independently (e.g. Main_tb-rule-engine-0, SequentialByOriginator_tb-rule-engine-1), so you can pinpoint which queue or partition is under load or accumulating errors.

Realtime chart. Shows the same per-partition breakdown for timeout-related outcomes:

Series	What it counts
Permanent Timeouts	Messages that timed out and exhausted all retries
Processing Timeouts	Messages that exceeded the processing timeout threshold
Permanent Failures	Cross-referenced with the Queue Stats panel

A flat line across all partitions means no failures or timeouts — expected during normal operation.

Table with a tab per queue partition. Each row is an exception logged during message processing:

Column	Description
Timestamp	When the exception occurred
Rule Chain	The rule chain where the exception was raised
Rule Node	The specific node that threw the exception
Latest Error	The error message or stack trace
Queue name	The queue the message was in
Service Id	The ThingsBoard service instance that processed the message

Use the Exceptions table to identify recurring node failures — the Rule Chain and Rule Node columns narrow the problem to an exact location in your processing graph.

Debug mode records incoming and outgoing messages at a specific rule node so you can inspect exactly what the node received, what it produced, and why it failed.

Open a node’s configuration window and click Debug. The Debug configuration dialog offers two independent modes:

Mode	Duration	What it records
Failures only	Always on (24/7)	Only messages that result in a Failure relation — low overhead, safe for production
All messages	15 minutes, then auto-disables	Every incoming and outgoing message with its full payload, metadata, and relation type

The default values below apply unless overridden by the tenant profile.

• DEBUG_SETTINGS_DEFAULT_DURATION_MINUTES (default: 15) — How long All messages mode stays active before auto-disabling, in minutes.
• ACTORS_RULE_CHAIN_DEBUG_MODE_RATE_LIMITS_PER_TENANT_ENABLED (default: true) — Enables the per-tenant cap on recorded debug events.
• ACTORS_RULE_CHAIN_DEBUG_MODE_RATE_LIMITS_PER_TENANT_CONFIGURATION (default: 50000:3600) — Maximum debug events per tenant per hour (count:seconds format).
• SQL_TTL_EVENTS_DEBUG_EVENTS_TTL (default: 604800) — How long debug events are retained in the database, in seconds (default = 1 week).

In the node’s configuration window, click the Events tab and select event type Debug. Each row shows one message passing through the node:

Column	Description
Event time	Timestamp of the processing event
Server	ThingsBoard service instance that processed the message
Type	IN (message entering the node) or OUT (message leaving the node)
Entity type	Originator entity type (e.g. DEVICE, ASSET)
Entity Id	Originator entity UUID
Message Id	Unique ID of the rule engine message
Message Type	Type string — see Message Types
Relation Type	Success, Failure, True, False, or custom — empty for IN events
Data	Full message payload JSON
Metadata	Message metadata key-value pairs
Error	Error details if the event resulted in a failure

Alternatively, click See debug events in the Debug configuration dialog to jump directly to the Events tab.

1. Open the Statistics dashboard and identify the Rule Chain and Rule Node in the Exceptions table.
2. Open that rule chain, find the node, and open its configuration window.
3. Enable Failures only debug mode and click Apply.
4. Wait for the next failure or replay a test message.
5. Click Events → Debug to inspect the IN and OUT rows — check Data, Metadata, and Error for the root cause.
6. Fix the node configuration or script, then disable All messages mode if it was enabled.