Scenario 2b of docs/soak-scenarios.md: 66 minutes of a steady 60 requests/s of lookups from a machine on the cluster’s network while one pod, two pods, all pods were replaced and the deployment was scaled to 16 and back.
The image was tag 10142026-4446caf (commit 4446caf, the merge of the graceful-shutdown change). Analysis: docs/performance-report.md, section 4.
summary.md, windows.csv, meta.json: the soak’s own output (66 one-minute windows).events.txt: the marks made while changing the deployment; events.md: what soak_analyze.py events made of them.watch.csv: the pods seen from the cluster side every 5 s (perf_tests/pod_watch.py).pods.jsonl (per-pod detail, 4.6 MB), canaries.jsonl and windows.jsonl; the maintainers keep them.