Kong Gateway OpenTelemetry metrics reference

In Kong Gateway, metrics are natively supported by the OpenTelemetry plugin. You can send metrics using the parameters under config.metrics.

To collect AI OTel metrics for AI Gateway, you will need to use the OpenTelemetry plugin with one of the AI Proxy or AI Proxy Advanced plugins. See Gen AI OpenTelemetry metrics reference for details.

The following metrics are exposed:

http.server.request.count

Total number of incoming HTTP requests.

• Instrument unit: {request}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.response.source	Origin of the current response. Possible values: upstream if the response originated by successfully contacting the upstream service kong otherwise
  kong.workspace.name	Name of the Workspace.
  http.request.method	Method used in the HTTP request.
  kong.response.status_code	HTTP status code of the response.

kong.latency.total

Complete end-to-end duration of a request in seconds.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.

kong.latency.internal

Kong’s internal processing time in seconds, from when the Gateway receives the request from the client to when it sends the request to the upstream service.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.

kong.latency.upstream

Upstream processing time in seconds, from when the Gateway sends the request to the upstream, to when the data is returned to Kong.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.

http.server.request.size

Size of each incoming HTTP request in bytes.

• Instrument unit: By
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.workspace.name	Name of the Workspace.

http.server.response.size

Total size of the HTTP response sent back to the client in bytes.

• Instrument unit: By
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.workspace.name	Name of the Workspace.

kong.shared_dict.usage

Current memory usage of a shared dict in bytes.

• Instrument unit: By
• Instrument type: Gauge

• Attributes:

  Attribute	Attribute description
  kong.shared_dict.name	Name of the shared dict.
  kong.subsystem	Nginx subsystem that produced the metric. Possible values: http stream

kong.shared_dict.size

Total memory size of a shared dict in bytes.

• Instrument unit: By
• Instrument type: Gauge

• Attributes:

  Attribute	Attribute description
  kong.shared_dict.name	Name of the shared dict.
  kong.subsystem	Nginx subsystem that produced the metric. Possible values: http stream

kong.memory.workers.lua_vm

Memory used by the worker’s Lua VM in bytes.

• Instrument unit: By
• Instrument type: Gauge

• Attributes:

  Attribute	Attribute description
  kong.pid	Worker process ID.
  kong.subsystem	Nginx subsystem that produced the metric. Possible values: http stream

kong.nginx.connection.count

Number of client connections in Nginx.

• Instrument unit: {connection}
• Instrument type: Gauge

• Attributes:

  Attribute	Attribute description
  kong.subsystem	Nginx subsystem that produced the metric. Possible values: http stream
  kong.connection.state	State of the client connection. Possible values: accepted handled total active reading writing waiting

kong.nginx.timer.count

Number of internal scheduled timers Nginx is running in the background.

• Instrument unit: {timer}
• Instrument type: Gauge

• Attributes:

  Attribute	Attribute description
  kong.timer.state	State of the timer. Possible values: pending running

kong.db.connection.status

Shows whether Kong has an active database connection. A value of 1 means connected. A value of 0 means not connected.

• Instrument unit: 1
• Instrument type: Gauge
• No attributes

kong.cp.connection.status

Shows whether the data plane has an active connection to the control plane. A value of 1 means connected. A value of 0 means not connected.

• Instrument unit: 1
• Instrument type: Gauge
• No attributes

kong.upstream.target.status

Upstream target’s health. The actual status is in the state attribute, with the metric value set to 1 when a state is populated.

• Instrument unit: 1
• Instrument type: Gauge

• Attributes:

  Attribute	Attribute description
  kong.upstream.name	Name of the Upstream.
  kong.target.address	Address of the Target.
  server.address	Address of the server.
  kong.upstream.state	Health of the Upstream Target. Possible values: healthy unhealthy dns_error
  kong.subsystem	Nginx subsystem that produced the metric. Possible values: http stream

kong.dp.cluster_cert.expiry

Timestamp when the data plane’s cluster certificate will expire.

• Instrument unit: s
• Instrument type: Gauge
• No attributes

kong.db.entity.count

Number of entities stored in Kong’s database.

• Instrument unit: {entity}
• Instrument type: Gauge
• No attributes

kong.db.entity.error.count

Number of errors seen during database entity count collection.

• Instrument unit: {error}
• Instrument type: Sum
• No attributes

kong.ee.license.signature

Last 8 bytes of the Enterprise license signature as a number.

• Instrument type: Gauge
• No attributes

kong.ee.license.expiration

Unix epoch time when the license expires, shifted by 24 hours to avoid timezone differences.

• Instrument unit: s
• Instrument type: Gauge
• No attributes

kong.ee.license.features

Indicates whether the data plane can read or write entities under the current license. Each capability (ee_entity_read and ee_entity_write) is reported as its own metric, where 1 means allowed and 0 means not allowed.

• Instrument unit: 1
• Instrument type: Gauge

• Attributes:

  Attribute	Attribute description
  kong.ee.license.feature	Enterprise feature. Possible values: ee_entity_read ee_entity_write

kong.ee.license.error.count

Number of errors that occurred while collecting license information.

• Instrument unit: {error}
• Instrument type: Sum
• No attributes

kong.websocket.server.total_connections v3.14+

Total number of incoming WebSocket connections, including active and previously closed connections.

• Instrument unit: {request}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.

kong.websocket.server.active_connections v3.14+

Number of currently active WebSocket connections.

• Instrument unit: {request}
• Instrument type: Gauge

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.

kong.websocket.server.received.message.size v3.14+

Size of incoming WebSocket message bodies in bytes, from the client.

• Instrument unit: By
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.

kong.websocket.server.sent.message.size v3.14+

Size of WebSocket messages sent back to the client in bytes.

• Instrument unit: By
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.

kong.websocket.handshake.error.count v3.14+

Number of errors seen during the WebSocket handshake.

• Instrument unit: {error}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  error.type	Type of error that occurred.

kong.websocket.connection.duration v3.14+

Duration of successfully established WebSocket connections in seconds.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.

gen_ai.client.operation.duration v3.14+

Total time Kong spends processing a Gen AI operation, such as an LLM request. Requires enable_request_metrics to populate the error.type attribute.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.
  error.type	Type of error that occurred.

gen_ai.server.request.duration v3.14+

Time the LLM provider spends processing the request. Requires enable_latency_metrics set to true. Requires enable_request_metrics to populate the error.type attribute.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.
  error.type	Type of error that occurred.

gen_ai.client.token.usage v3.14+

Number of tokens consumed by the Gen AI operation.

• Instrument unit: {token}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.token.type	Token category: input, output, or total.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.

gen_ai.server.time_to_first_token v3.14+

Time from when the model server receives the request until the first output token is generated.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.

gen_ai.server.time_per_output_token v3.14+

Time between successive output tokens generated by the model server after the first token.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.

kong.gen_ai.llm.cost v3.14+

Cost of AI requests.

• Instrument unit: {cost}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.gen_ai.cache.status	Cache status: hit or empty if not cached.
  kong.gen_ai.vector_db	Vector database used for caching, such as redis.
  kong.gen_ai.embeddings.provider	Embeddings provider used for caching.
  kong.gen_ai.embeddings.model	Embeddings model used for caching.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.

kong.gen_ai.cache.fetch.latency v3.14+

Time to fetch a response from the semantic cache.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.gen_ai.cache.status	Cache status: hit or empty if not cached.
  kong.gen_ai.vector_db	Vector database used for caching, such as redis.
  kong.gen_ai.embeddings.provider	Embeddings provider used for caching.
  kong.gen_ai.embeddings.model	Embeddings model used for caching.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.

kong.gen_ai.cache.embeddings.latency v3.14+

Time to generate embeddings during cache operations.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.gen_ai.cache.status	Cache status: hit or empty if not cached.
  kong.gen_ai.vector_db	Vector database used for caching, such as redis.
  kong.gen_ai.embeddings.provider	Embeddings provider used for caching.
  kong.gen_ai.embeddings.model	Embeddings model used for caching.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.

kong.gen_ai.rag.fetch.latency v3.14+

Time to fetch data from a RAG source.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.gen_ai.cache.status	Cache status: hit or empty if not cached.
  kong.gen_ai.vector_db	Vector database used for caching, such as redis.
  kong.gen_ai.embeddings.provider	Embeddings provider used for caching.
  kong.gen_ai.embeddings.model	Embeddings model used for caching.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.

kong.gen_ai.rag.embeddings.latency v3.14+

Time to generate embeddings for RAG operations.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  gen_ai.provider.name	Name of the Gen AI provider.
  gen_ai.request.model	Model name targeted by the request.
  gen_ai.response.model	Model name reported by the provider in the response.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.
  kong.gen_ai.cache.status	Cache status: hit or empty if not cached.
  kong.gen_ai.vector_db	Vector database used for caching, such as redis.
  kong.gen_ai.embeddings.provider	Embeddings provider used for caching.
  kong.gen_ai.embeddings.model	Embeddings model used for caching.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.
  kong.gen_ai.request.mode	Request mode: oneshot, stream, or realtime.

kong.gen_ai.aws.guardrails.latency v3.14+

Time for AWS Guardrails to process a request.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.gen_ai.aws.guardrails.id	ID of the AWS Guardrails configuration.
  kong.gen_ai.aws.guardrails.version	Version of the AWS Guardrails configuration.
  kong.gen_ai.aws.guardrails.mode	Mode of the AWS Guardrails evaluation.
  kong.gen_ai.aws.guardrails.region	AWS region of the Guardrails service.
  kong.workspace.name	Name of the Workspace.
  kong.auth.consumer.name	Name of the authenticated Consumer.

kong.gen_ai.mcp.response.size v3.14+

Size of the MCP response body in bytes.

• Instrument unit: By
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  mcp.method.name	MCP method name, such as tools/call.
  gen_ai.tool.name	Name of the MCP tool invoked.

kong.gen_ai.mcp.request.error.count v3.14+

Number of MCP request errors.

• Instrument unit: {error}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  mcp.method.name	MCP method name, such as tools/call.
  gen_ai.tool.name	Name of the MCP tool invoked.
  error.type	Type of error that occurred.

mcp.client.operation.duration v3.14+

Duration of the MCP request as observed by the sender. Only available when the AI MCP Proxy plugin is in passthrough-listener mode. Requires enable_latency_metrics set to true.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  mcp.method.name	MCP method name, such as tools/call.
  gen_ai.tool.name	Name of the MCP tool invoked.
  error.type	Type of error that occurred.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.

mcp.server.operation.duration v3.14+

Duration of the MCP request as observed by the receiver.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  mcp.method.name	MCP method name, such as tools/call.
  gen_ai.tool.name	Name of the MCP tool invoked.
  error.type	Type of error that occurred.
  gen_ai.operation.name	Operation requested, such as chat or embeddings.

kong.gen_ai.mcp.acl.allowed v3.14+

Number of MCP requests allowed by ACL rules.

• Instrument unit: {request}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.gen_ai.mcp.primitive	MCP primitive type, such as tool.
  kong.gen_ai.mcp.primitive_name	Name of the MCP primitive.

kong.gen_ai.mcp.acl.denied v3.14+

Number of MCP requests denied by ACL rules.

• Instrument unit: {request}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.gen_ai.mcp.primitive	MCP primitive type, such as tool.
  kong.gen_ai.mcp.primitive_name	Name of the MCP primitive.

kong.gen_ai.a2a.request.count v3.14+

Total number of A2A requests.

• Instrument unit: {request}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.gen_ai.a2a.method	A2A method name.
  kong.gen_ai.a2a.binding	A2A binding type.

kong.gen_ai.a2a.request.duration v3.14+

Duration of an A2A request in seconds.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.gen_ai.a2a.method	A2A method name.
  kong.gen_ai.a2a.binding	A2A binding type.

kong.gen_ai.a2a.response.size v3.14+

Size of the A2A response body in bytes.

• Instrument unit: By
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.gen_ai.a2a.method	A2A method name.
  kong.gen_ai.a2a.binding	A2A binding type.

kong.gen_ai.a2a.ttfb v3.14+

Time to first byte for A2A streaming responses in seconds.

• Instrument unit: s
• Instrument type: Histogram

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.gen_ai.a2a.method	A2A method name.
  kong.gen_ai.a2a.binding	A2A binding type.

kong.gen_ai.a2a.request.error.count v3.14+

Number of A2A request errors.

• Instrument unit: {error}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.gen_ai.a2a.method	A2A method name.
  kong.gen_ai.a2a.binding	A2A binding type.
  kong.gen_ai.a2a.error.type	Type of the A2A error.

kong.gen_ai.a2a.task.state.count v3.14+

Number of A2A task state transitions.

• Instrument unit: {state}
• Instrument type: Sum

• Attributes:

  Attribute	Attribute description
  kong.service.name	Name of the Gateway Service.
  kong.route.name	Name of the Route.
  kong.workspace.name	Name of the Workspace.
  kong.gen_ai.a2a.task.state	Task state, such as completed, failed, or in_progress.