How ragull-11/clusterloader works
ClusterLoader is a Kubernetes workload benchmarking harness: operators select a workload configuration, run CL2 while collecting cluster and Prometheus metrics, and save tracker/report outputs. Its workload scenarios cover pod scale, churn, deployment scaling, jobs, and parallel create/delete. A separate backfill utility repairs missing Prometheus metrics in saved tracker columns. The README only names the product, and the runner excerpt is partial, so unshown runtime wiring is omitted rather than inferred.
Main entry points: cl2_runner.py CL2 runner · backfill_prometheus.py Metric backfill
Components 14
Benchmark workloads
- Pod scale
configs/testcase-1-high-pod-count - Pod churn
configs/testcase-2-pod-churn - Deployment scaling
configs/testcase-3-deployment-scaling - Jobs
configs/testcase-4-jobs - Parallel create/delete
configs/testcase-5-parallel-create-delete - Resource templates
configs/templates
Run and observe
- CL2 runner
scripts/cl2_runner.py - Pod watcher
scripts/cl2_runner.py
Metrics and results
- Prometheus
- CSV tracker
scripts/cl2_runner.py - Run reports
scripts/cl2_runner.py - Metric backfill
scripts/backfill_prometheus.py - Formatted workbook
scripts/cl2_runner.py
Other components
- Operator
Connections 9
- Operator → CL2 runner: starts run
scripts/cl2_runner.py - CL2 runner → Pod watcher: starts watcher
scripts/cl2_runner.py - CL2 runner → Prometheus: queries metrics
- CL2 runner → CSV tracker: writes metrics
scripts/cl2_runner.py - CL2 runner → Run reports: writes reports
scripts/cl2_runner.py - Metric backfill → CL2 runner: imports and calls
scripts/backfill_prometheus.py - Metric backfill → Prometheus: requeries metrics
scripts/backfill_prometheus.py - Metric backfill → CSV tracker: patches columns
scripts/backfill_prometheus.py - Metric backfill → Formatted workbook: rebuilds workbook
scripts/backfill_prometheus.py