Skip to content

README.rst

Latest commit

 

History

History
67 lines (49 loc) · 2.32 KB

README.rst

File metadata and controls

67 lines (49 loc) · 2.32 KB

Multi-Destination Exporting

This example shows how to export telemetry data to multiple destinations simultaneously. As described in the OTLP specification, each destination should have implemented its own queuing, acknowledgement handling, and retry logic to prevent one slow or unavailable destination from blocking the others.

The OpenTelemetry Python SDK achieves this by using a separate processor or reader per destination:

  • Traces: Use one BatchSpanProcessor per destination, each wrapping its own SpanExporter. Add each processor to the TracerProvider via add_span_processor().
  • Metrics: Pass multiple MetricReader instances (each wrapping its own MetricExporter) to the MeterProvider constructor via the metric_readers parameter.
  • Logs: Use one BatchLogRecordProcessor per destination, each wrapping its own LogExporter. Add each processor to the LoggerProvider via add_log_record_processor().

Note

The Profiles signal is not yet supported in the Python SDK. When it becomes available, the same pattern will apply.

The source files of these examples are available :scm_web:`here <docs/examples/multi-destination-exporting/>`.

Installation

pip install opentelemetry-api
pip install opentelemetry-sdk
pip install opentelemetry-exporter-otlp-proto-grpc
pip install opentelemetry-exporter-otlp-proto-http
pip install opentelemetry-instrumentation-logging # For LoggingHandler

Run the Example

python multi_destination_traces.py
python multi_destination_metrics.py
python multi_destination_logs.py

The output will be shown in the console for the ConsoleSpanExporter, ConsoleMetricExporter, and ConsoleLogRecordExporter destinations. The OTLP destinations require a running collector.

Useful links