Skip to main content

Examples

Real-world examples demonstrating how to use Konductor synchronization primitives to solve common coordination problems in Kubernetes.

Example Categories​

ETL and Data Processing​

Database Operations​

CI/CD and Deployments​

Microservices Patterns​

Quick Reference​

Common Patterns​

InitContainer Gating​

initContainers:
- name: wait-dependencies
  image: logiciq/koncli:latest
  command:
  - koncli
  - barrier
  - wait
  - dependencies-ready

Semaphore in Script​

if koncli semaphore acquire resource-limit --ttl=10m; then
  # Do work
  koncli semaphore release resource-limit
fi

Singleton CronJob​

HOLDER_ID="$HOSTNAME-$$"
if koncli lease acquire singleton-job --holder $HOLDER_ID --timeout=0; then
  # Run job
  koncli lease release singleton-job --holder $HOLDER_ID
fi

Pipeline Stage Coordination​

# Stage 1: Signal completion
koncli barrier arrive stage-1-complete

# Stage 2: Wait for stage 1
koncli barrier wait stage-1-complete --timeout=30m

Integration Examples​

Argo Workflows​

apiVersion: argoproj.io/v1alpha1
kind: Workflow
metadata:
  name: coordinated-workflow
spec:
  templates:
  - name: wait-barrier
    container:
      image: logiciq/koncli:latest
      command: [koncli, barrier, wait, "{{inputs.parameters.barrier-name}}"]

Tekton Pipelines​

apiVersion: tekton.dev/v1beta1
kind: Task
metadata:
  name: acquire-semaphore
spec:
  steps:
  - name: acquire
    image: logiciq/koncli:latest
    script: |
      koncli semaphore acquire "$(params.semaphore-name)" --ttl="$(params.ttl)"

Flux/GitOps​

apiVersion: kustomize.toolkit.fluxcd.io/v1beta2
kind: Kustomization
metadata:
  name: app-deployment
spec:
  dependsOn:
  - name: database-migration
  postBuild:
    substitute:
      BARRIER_NAME: "deployment-ready"

Performance Examples​

High-Throughput Processing​

# Process 1000 items with max 50 concurrent workers
apiVersion: konductor.io/v1
kind: Semaphore
metadata:
  name: worker-pool
spec:
  permits: 50
  ttl: 5m

Large-Scale Coordination​

# Coordinate 100 parallel jobs
apiVersion: konductor.io/v1
kind: Barrier
metadata:
  name: massive-parallel
spec:
  expected: 100
  timeout: 2h
  quorum: 95  # Allow 5% failure rate

Troubleshooting Examples​

Debug Stuck Barriers​

# Check barrier status
kubectl describe barrier my-barrier

# List arrivals
kubectl get barrier my-barrier -o jsonpath='{.status.arrivals}'

# Check for missing jobs
kubectl get jobs -l barrier=my-barrier

Monitor Semaphore Usage​

# Watch semaphore status
watch kubectl get semaphore my-semaphore

# Check permit holders
kubectl get semaphore my-semaphore -o jsonpath='{.status.holders}'

Lease Debugging​

# Check lease holder
kubectl get lease my-lease -o jsonpath='{.status.holder}'

# Verify holder is alive
kubectl get pods | grep "$(kubectl get lease my-lease -o jsonpath='{.status.holder}')"

Best Practices from Examples​

  1. Use meaningful names - Choose descriptive resource names
  2. Set appropriate timeouts - Balance safety with responsiveness
  3. Handle failures gracefully - Always clean up resources
  4. Monitor coordination state - Use status fields for observability
  5. Test coordination logic - Verify behavior under failure conditions

Contributing Examples​

Have a useful Konductor pattern? We'd love to include it!

  1. Fork the repository
  2. Add your example to the appropriate category
  3. Include complete YAML manifests and explanations
  4. Submit a pull request

Next Steps​

  • Choose an example that matches your use case
  • Adapt the patterns to your specific requirements
  • Check the API Reference for detailed configuration options
  • Use the CLI Reference for command-line integration