Installing Jaeger v2: all-in-one, configuration and storage

1. All-in-one #

For learning and testing.

Startup #

docker run --rm --name jaeger \
  -p 16686:16686 -p 4317:4317 -p 4318:4318 \
  jaegertracing/jaeger:latest
# Interface : http://localhost:16686

Common mistake: The all-in-one stores traces in memory: everything disappears on restart. Use Badger (local disk) for a persistent test environment, and distributed storage in production.

2. The configuration file #

OpenTelemetry Collector syntax.

Structure #

service:
  extensions: [jaeger_storage, jaeger_query, healthcheckv2]
  pipelines:
    traces:
      receivers: [otlp]
      processors: [batch]
      exporters: [jaeger_storage_exporter]

extensions:
  healthcheckv2:
    use_v2: true
    http:
  jaeger_query:
    storage:
      traces: stockage_principal
  jaeger_storage:
    backends:
      stockage_principal:
        memory:
          max_traces: 100000

receivers:
  otlp:
    protocols:
      grpc: { endpoint: 0.0.0.0:4317 }
      http: { endpoint: 0.0.0.0:4318 }

processors:
  batch:

exporters:
  jaeger_storage_exporter:
    trace_storage: stockage_principal

docker run --rm -v $PWD/config.yaml:/jaeger/config.yaml -p 16686:16686 -p 4317:4317 -p 4318:4318 \
  jaegertracing/jaeger:latest --config /jaeger/config.yaml

Extensions declare the storage backends (jaeger_storage) and the interface (jaeger_query); the pipeline receives OTLP and writes via jaeger_storage_exporter. Any Collector processor (filter, attributes, sampling) can be inserted into the pipeline.

Good to know: Key names change between Jaeger v2 versions: start from the example files in the repository (cmd/jaeger/*.yaml) matching your version.

3. Storage backends #

Choosing by volume.

Options #

extensions:
  jaeger_storage:
    backends:
      es_traces:
        elasticsearch:
          server_urls: [http://elasticsearch:9200]
          indices:
            index_prefix: jaeger
            spans: { date_layout: "2006-01-02", rollover_frequency: day }

Good practice: With Elasticsearch or OpenSearch, retention is managed by deleting old indices: Index Lifecycle Management (ILM) or the jaeger-es-index-cleaner job.