Install the HAProxy OpenTelemetry filter

To begin collecting OpenTelemetry data from your load balancer, you must install the filter. This page describe the steps.

It also explains how to run OpenTelemetry Collector to test your installation. OpenTelemetry Collector is suitable for test and production environments, although in production you’ll want to run it on a separate server from your load balancer. It can print telemetry data to the console so you can quickly verify that everything works. Or use it to relay data to other software such as Jaeger and Prometheus, which let you visualize traces, metrics, and logs.

Install the OpenTelemetry filter Jump to heading #

Installing the HAProxy OpenTelemetry filter varies depending on the product. The table below tells you where to go.

Once installed, you can run OpenTelemetry Collector to verify that everything works. Also, check out the tutorials to learn how to configure traces, logs, and metrics.

Run OpenTelemetry Collector Jump to heading #

OpenTelemetry Collector provides a vendor-agnostic way to receive, process, and export telemetry data. You can send your traces, logs, and metrics to it to see the data as it streams in. Simply configure its debug exporter, which prints to the console. Or, configure Collector to relay data to any software that supports the OpenTelemetry protocol, including Elasticsearch, Jaeger, Prometheus, and others. For a full list, see the OpenTelemetry Registry.

In this tutorial, we’ll run OpenTelemetry Collector by using Docker Compose.

1. Install Docker and Docker Compose by following the official guide, Install Docker Engine.

2. Create a Docker Compose file that runs OpenTelemetry Collector as a container. In this example, we expose it at port 4318. Also, we run Jaeger and Prometheus as containers to provide destinations for traces and metrics.

   docker-compose.yaml
   yaml
   networks:
     custom_net:
       driver: bridge
       ipam:
         config:
           - subnet: 172.16.0.0/16
             gateway: 172.16.0.1
   services:
     otel-collector:
       image: otel/opentelemetry-collector
       volumes:
         - ./collector-config.yaml:/etc/otelcol/config.yaml
       ports:
         - 4318:4318 # OTLP http receiver
       networks:
         custom_net:
           ipv4_address: 172.16.0.12
     # Send traces to Jaeger at 172.16.0.13:4318
     jaeger:
       image: jaegertracing/jaeger
       ports:
         - "16686:16686"
         - "16687:16687"
       environment:
         - LOG_LEVEL=debug
       networks:
         custom_net:
           ipv4_address: 172.16.0.13
      
     # Send metrics to Prometheus at 172.16.0.14:9090
     prometheus:
       image: prom/prometheus
       ports:
         - "9090:9090"
       volumes:
         - "./prometheus.yaml:/etc/prometheus/prometheus.yml"
       networks:
         custom_net:
           ipv4_address: 172.16.0.14

   docker-compose.yaml
   yaml
   networks:
     custom_net:
       driver: bridge
       ipam:
         config:
           - subnet: 172.16.0.0/16
             gateway: 172.16.0.1
   services:
     otel-collector:
       image: otel/opentelemetry-collector
       volumes:
         - ./collector-config.yaml:/etc/otelcol/config.yaml
       ports:
         - 4318:4318 # OTLP http receiver
       networks:
         custom_net:
           ipv4_address: 172.16.0.12
     # Send traces to Jaeger at 172.16.0.13:4318
     jaeger:
       image: jaegertracing/jaeger
       ports:
         - "16686:16686"
         - "16687:16687"
       environment:
         - LOG_LEVEL=debug
       networks:
         custom_net:
           ipv4_address: 172.16.0.13
      
     # Send metrics to Prometheus at 172.16.0.14:9090
     prometheus:
       image: prom/prometheus
       ports:
         - "9090:9090"
       volumes:
         - "./prometheus.yaml:/etc/prometheus/prometheus.yml"
       networks:
         custom_net:
           ipv4_address: 172.16.0.14

3. Create configuration files in the same directory as the docker-compose.yaml file.

   • For OpenTelemetry Collector, create a file named collector-config.yaml.

     collector-config.yaml
     yaml
     receivers:
       otlp:
         protocols:
           http:
             endpoint: 0.0.0.0:4318
     exporters:
       debug:
         verbosity: detailed
       otlp_http:
         # The base URL of your OTLP/HTTP compatible backend (e.g., Jaeger)
         endpoint: "http://172.16.0.13:4318"
       prometheus:
         # publish Prometheus metrics to be scraped
         endpoint: "0.0.0.0:9090"
     service:
       pipelines:
         traces:
           receivers: [otlp]
           exporters: [debug, otlp_http]
         metrics:
           receivers: [otlp]
           exporters: [debug, prometheus]
         logs:
           receivers: [otlp]
           exporters: [debug]

     collector-config.yaml
     yaml
     receivers:
       otlp:
         protocols:
           http:
             endpoint: 0.0.0.0:4318
     exporters:
       debug:
         verbosity: detailed
       otlp_http:
         # The base URL of your OTLP/HTTP compatible backend (e.g., Jaeger)
         endpoint: "http://172.16.0.13:4318"
       prometheus:
         # publish Prometheus metrics to be scraped
         endpoint: "0.0.0.0:9090"
     service:
       pipelines:
         traces:
           receivers: [otlp]
           exporters: [debug, otlp_http]
         metrics:
           receivers: [otlp]
           exporters: [debug, prometheus]
         logs:
           receivers: [otlp]
           exporters: [debug]

   • For Prometheus, create a file named prometheus.yaml.

     prometheus.yaml
     yaml
     # my global config
     global:
       scrape_interval: 15s # Scrape every 15 seconds. Default is every 1 minute.
     scrape_configs:
       - job_name: "prometheus"
         static_configs:
           - targets: ["localhost:9090"]
       - job_name: "opentelemetry"
         scheme: http
         static_configs:
         - targets: ['172.16.0.12:9090']

     prometheus.yaml
     yaml
     # my global config
     global:
       scrape_interval: 15s # Scrape every 15 seconds. Default is every 1 minute.
     scrape_configs:
       - job_name: "prometheus"
         static_configs:
           - targets: ["localhost:9090"]
       - job_name: "opentelemetry"
         scheme: http
         static_configs:
         - targets: ['172.16.0.12:9090']

4. Start the containers.

   nix
   sudo docker compose up

   nix
   sudo docker compose up

   Once the containers are running, you can see the OpenTelemetry Collector’s debug log in the console; the Jaeger UI listens at HTTP port 16686 and the Prometheus UI listens at port 9090. You won’t see metrics until you add them to your instrumentation.cfg file.