Qubership Jaeger

Helm charts that deploy a production-ready Jaeger stack on Kubernetes/OpenShift, curated by the Qubership team.

Features / key capabilities

• One-click install of the full Jaeger stack (agent, collector, query, UI)
• Pluggable storage backends: Cassandra (default) or OpenSearch
• TLS everywhere – built-in issuers & secrets for mTLS between components
• Optional Envoy proxy with Basic or OAuth2 auth in front of the Query UI
• Grafana dashboards, Prometheus PodMonitor, readiness probes
• Robot-Framework integration tests & Hotrod demo application
• Memory limiter processor – Prevents OOM kills by managing memory usage and applying backpressure

Requirements

• Kubernetes ≥ 1.23 or OpenShift ≥ 4.10
• Helm ≥ 3.8
• For production: a Cassandra or OpenSearch cluster reachable from the namespace

Installation / Quick start

# Add the public repo hosted on GitHub Pages
helm repo add qubership-jaeger https://netcracker.github.io/qubership-jaeger
helm repo update

# Install with the default values
helm install jaeger qubership-jaeger/qubership-jaeger \
  --namespace jaeger --create-namespace

Use --version <chart_version> to pin a specific release.

Usage

• Access the Jaeger UI (port-forward or via the provided ingress):

  kubectl port-forward svc/jaeger-query 16686 -n jaeger
  open http://localhost:16686

• Send spans in Zipkin format to jaeger-collector:9411 or native gRPC/HTTP to jaeger-collector service.

• Hotrod demo can be enabled via hotrod.install=true.

Configuration

Extensive configuration options are documented in the Configuration section of the docs. Override any parameter from values.yaml via --set or a custom values.yaml.

Examples:

# Enable OpenSearch storage
helm upgrade --install jaeger qubership-jaeger/qubership-jaeger \
  -n jaeger -f - <<EOF
jaeger:
  storage:
    type: opensearch
EOF

# Enable memory limiter to prevent OOM issues
helm upgrade --install jaeger qubership-jaeger/qubership-jaeger \
  -n jaeger --reuse-values \
  --set collector.config.processors.memory_limiter.enabled=true \
  --set collector.config.processors.memory_limiter.limit_mib=2048 \
  --set collector.config.processors.memory_limiter.spike_limit_mib=256

Note: The memory limiter should be configured with limit_mib set to approximately 75-80% of the pod's memory limit to provide a safety margin.

API Reference / Documentation

The full documentation is published at https://netcracker.github.io/qubership-jaeger/ and mirrors the Markdown files under /docs.

Architecture / Diagram

System overview

flowchart TD
    Client["Application / Hotrod"] --> Agent(Agent)
    Agent --> Collector(Collector)
    Collector -->|"Store spans"| Storage[(Cassandra / OpenSearch)]
    Query(Query) <-->|"Read spans"| Storage
    UI["Jaeger UI"] --> Query

Loading

Testing

Integration and smoke tests can be run against a live cluster:

helm upgrade --install jaeger-tests qubership-jaeger/qubership-jaeger \
  -n jaeger -f - <<EOF
integrationTests:
  install: true
  tags: smokeORha
  image: ghcr.io/netcracker/jaeger-integration-tests:main
EOF

The Robot-Framework suites live under integration-tests/robot and are executed inside the test runner Pod.

Contributing

Please read CONTRIBUTING.md before opening a PR. Run the Super-Linter locally:

docker run \
  -e RUN_LOCAL=true \
  -e DEFAULT_BRANCH=$(git rev-parse --abbrev-ref HEAD) \
  --env-file .github/super-linter.env \
  -v ${PWD}:/tmp/lint \
  --rm \
  ghcr.io/super-linter/super-linter:slim-$(sed -nE 's#.*uses:\s+super-linter/super-linter/slim@([^\s]+).*#\1#p' .github/workflows/super-linter.yaml)

License

This project is licensed under the terms of the Apache 2.0 License.