You are browsing documentation for a version of Kuma that is not the latest release.

Looking for even older versions? Learn more.

Traffic Log

With TrafficLog policy you can easily set up access logs on every data-plane in a Mesh.

TrafficLog only logs outbound traffic. It doesn’t log inbound traffic.

Configuring access logs in Kuma is a 3-step process:

Add a logging backend

A logging backend is essentially a sink for access logs.

Currently, a logging backend can be either a file or a TCP log collector, such as Logstash.

kind: Mesh
  name: default
    # TrafficLog policies may leave the `backend` field undefined.
    # In that case the logs will be forwarded into the `defaultBackend` of that Mesh.
    defaultBackend: file
    # List of logging backends that can be referred to by name
    # from TrafficLog policies of that Mesh.
      - name: logstash
        # Use `format` field to adjust the access log format to your use case.
        format: '{"start_time": "%START_TIME%", "source": "%KUMA_SOURCE_SERVICE%", "destination": "%KUMA_DESTINATION_SERVICE%", "source_address": "%KUMA_SOURCE_ADDRESS_WITHOUT_PORT%", "destination_address": "%UPSTREAM_HOST%", "duration_millis": "%DURATION%", "bytes_received": "%BYTES_RECEIVED%", "bytes_sent": "%BYTES_SENT%"}'
        type: tcp
        # Use `config` field to co configure a TCP logging backend.
          # Address of a log collector.
      - name: file
        type: file
        # Use `file` field to configure a file-based logging backend.
          path: /tmp/access.log
        # When `format` field is omitted, the default access log format will be used.

Add a TrafficLog resource

You need to create a TrafficLog policy to select a subset of traffic and forward its access logs into one of the logging backends configured for that Mesh.

kind: TrafficLog
  name: all-traffic
mesh: default
  # This TrafficLog policy applies all traffic in that Mesh.
    - match: "*"
    - match: "*"
  # When `backend ` field is omitted, the logs will be forwarded into the `defaultBackend` of that Mesh.
kind: TrafficLog
  name: backend-to-database-traffic
mesh: default
  # This TrafficLog policy applies only to traffic from service `backend` to service `database`.
    - match: backend_kuma-example_svc_8080
    - match: database_kuma-example_svc_5432
    # Forward the logs into the logging backend named `logstash`.
    backend: logstash

When backend field of a TrafficLog policy is omitted, the logs will be forwarded into the defaultBackend of that Mesh.

Log aggregation and visualisation

Kuma is presenting a simple solution to aggregate the logs of your containers and the access logs of your data-planes.

1. Install Loki

To install Loki use kumactl install logging | kubectl apply -f -. This will deploy Loki automatically in a kuma-logging namespace.

2. Update the mesh

The logging backend needs to be configured to send the access logs of your data-planes to stdout. Loki will directly retrieve the logs from stdout of your containers.

kind: Mesh
  name: default
      defaultBackend: loki
        - name: loki
          type: file
            path: /dev/stdout

3. Configure Grafana to visualize the logs

To visualise your containers’ logs and your access logs you need to have a Grafana up and running. You can install Grafana by following the information of the official page or use the one installed with Traffic metrics.

If you have installed Grafana yourself you can configure a new datasource with url:http://loki.kuma-logging:3100 so Grafana will be able to retrieve the logs from Loki.

Loki Grafana configuration

At this point you can visualize your containers’ logs and your access logs in Grafana by choosing the loki datasource in the explore section.

For example, running: {container="kuma-sidecar"} |= "GET" will show all GET requests on your cluster. To learn more about the search syntax check the Loki docs.

Nice to have

Having your Logs and Traces in the same visualisation tool can come really handy. By adding the traceId in your app logs you can visualize your logs and the related Jaeger traces. To learn more about it go read this article.

To set up tracing see the traffic-trace policy.

You can also forward the access logs to a collector (such as logstash) that can further transmit them into systems like Splunk, ELK and Datadog.

Access Log Format

Kuma gives you full control over the format of access logs.

The shape of a single log record is defined by a template string that uses command operators to extract and format data about a TCP connection or an HTTP request.



where %START_TIME% and %KUMA_SOURCE_SERVICE% are examples of available command operators.

A complete set of supported command operators consists of:

  1. All command operators supported by Envoy
  2. Command operators unique to Kuma

The latter include:

Command Operator Description
%KUMA_MESH% name of the mesh in which traffic is flowing
%KUMA_SOURCE_SERVICE% name of a service that is the source of traffic
%KUMA_DESTINATION_SERVICE% name of a service that is the destination of traffic
%KUMA_SOURCE_ADDRESS_WITHOUT_PORT% address of a Dataplane that is the source of traffic

Access Logs for TCP and HTTP traffic

All access log command operators are valid to use with both TCP and HTTP traffic.

If a command operator is specific to HTTP traffic, such as %REQ(X?Y):Z% or %RESP(X?Y):Z%, it will be replaced by a symbol “-” in case of TCP traffic.

Internally, Kuma determines traffic protocol based on the value of tag on the inbound interface of a destination Dataplane.

The default format string for TCP traffic is:


The default format string for HTTP traffic is:


To provide different format for TCP and HTTP logging you can define two separate logging backends with the same address and different format. Then define two TrafficLog entity, one for TCP and one for HTTP with http selector.

Access Logs in JSON format

If you need an access log with entries in JSON format, you have to provide a template string that is a valid JSON object, e.g.

  "start_time":          "%START_TIME%",
  "source":              "%KUMA_SOURCE_SERVICE%",
  "destination":         "%KUMA_DESTINATION_SERVICE%",
  "source_address":      "%KUMA_SOURCE_ADDRESS_WITHOUT_PORT%",
  "destination_address": "%UPSTREAM_HOST%",
  "duration_millis":     "%DURATION%",
  "bytes_received":      "%BYTES_RECEIVED%",
  "bytes_sent":          "%BYTES_SENT%"

To use it with Logstash, use json_lines codec and make sure your JSON is formatted into one line.

Logging external services

When running Kuma on Kubernetes you can also log the traffic to external services. To do it, the matched destination section has to have wildcard * value. In such case %KUMA_DESTINATION_SERVICE% will have value external and %UPSTREAM_HOST% will have an IP of the service.


TrafficLog is an Outbound Connection Policy. The only supported value for destinations.match is

Builtin Gateway support

Traffic Log is a Kuma outbound connection policy, so Kuma chooses a Traffic Log policy by matching the service tag of the Dataplane’s outbounds. Since a builtin gateway Dataplane does not have outbounds, Kuma always uses the builtin service name “pass_through” to match the Traffic Log policy for Gateways.

Last Updated: 6/23/2023, 11:00:36 AM