StackState curated integration
The StackState Agent V2 Traefik integration provides the following functionality:
- Reporting Traefik frontends and backends as topology elements.
- Reporting all network connections between services, including network traffic telemetry.
# Tracing definition
# Use the Datadog backend to send the Datadog Tracing format to StackState Agent
backend = "datadog"
# Component name used for your Traefik instance in StackState
serviceName = "traefik"
# Span name limit allows for name truncation in case of very long Frontend/Backend names
# This can prevent certain tracing providers to drop traces that exceed their length limits
spanNameLimit = 100
# StackState Agent Host Port instructs reporter to send spans to the StackState Agent at this address
localAgentHostPort = "agentHost:8126"
# Applies a shared tag in a form of source:traefik to all the spans of the trace
globalTag = "source:traefik"
When using Traefik in conjunction with one of our language specific trace clients, eg. StackState Java Trace Client - Java APM it is important to note that you should use the
backendname of your Traefik service as the
service-namefor the trace client to allow automatic merging of the service components within StackState.
Eg. for the following
backend = "stackstate-demo-backend"
rule = "Host:test.stackstate-demo-backend.localhost"
url = "..."
you should pass the following jvm argument when starting your java application:
or for a similar docker-compose configuration:
pid: "host" # use pid:"host" to ensure pid's match with processes reported by the StackState process agent
To verify whether the StackState Trace Agent has received traces, set the logging level to debug and check the
In Docker or Kubernetes, set the following environment variable for the StackState Agent