How It Works
- Your application exports OTLP metrics over gRPC to
localhost:4317 - The agent receives, transforms, and buffers the metrics
- Metrics are uploaded to Chamber in batches every 60 seconds
- They appear on the Services dashboard alongside infrastructure metrics
Configuration
The OTLP receiver is enabled by default on port 4317. No additional configuration is required for most setups.
To disable the receiver, set
CHAMBER_OTLP_GRPC_ENABLED=false in /etc/chamber/agent.env and restart the agent.
Sending Metrics from Your Application
Point your application’s OTLP exporter to the agent’s gRPC endpoint. Here is an example using the Python OpenTelemetry SDK:Set
service.name in your resource attributes — it appears as the service name label on all metrics in the Chamber dashboard.Metric Name Transformation
The agent transforms OTLP metric names to a Prometheus-compatible format before uploading:Unit Suffix Mapping
Histogram Support
OTLP histograms are decomposed into Prometheus-style metrics:Supported Metric Types
Verifying Metrics Are Flowing
After configuring your application, check the agent logs:service_name label you configured.
Next Steps
Metrics Reference
Full list of infrastructure metrics collected
Configuration
All environment variables and settings

