mirror of
https://github.com/windmill-labs/windmill.git
synced 2026-08-25 00:01:55 +00:00
74b662d8de
Stand up a minikube-backed simulation subsystem for benching Windmill under realistic multi-node load, with a per-bench measurement pipeline and a dashboard renderer that consolidates throughput, queue depth, per-node CPU, PG latency/conns, OOM events, and per-node CPU-util-vs-oversaturation into one SVG report. Sim infrastructure (sim/): - k8s_provisioner: minikube up + heterogeneous node sizing from topology JSON - helm_deploy: helm install Windmill with smoke.yaml + local.yaml overlays - image_cache: pre-load required images so bench bringup is offline-safe - toxiproxy_k8s: per-node toxiproxy DaemonSet for cross-node latency injection - cpu_sampler_k8s: privileged DS reading per-cgroup cpu.stat at 10Hz, dual- writes to stdout AND a host-mounted log file (/var/log/wm-sim-cpu-sampler/ sampler.tsv) so heavy benches no longer lose early samples to kubelet log rotation - pg_logging: ALTER SYSTEM + SIGHUP to enable verbose PG logging without restart - pgbadger: post-bench PG log analysis HTML report - readiness: pre-bench cluster health check (samplers stable ≥30s, workers ready, PG responsive, queue empty, **deploy.status rollout-complete**) — the rollout-complete check catches mid-rolling-update fires that previously starved m04's sampler under cgroup_mutex contention Per-bench JSONL pollers, started/finalized alongside the bench loop: - pod_timeline: 1Hz workers-per-node Ready counts (used for the workers panel) - oom_poller: live OOM event capture (kernel + kubelet evictions + cgroup) - pg_latency_poller: 4Hz psql \\timing on SELECT 1 vs kubectl-exec roundtrip - pg_conn_poller: 1Hz pg_stat_activity by state (active/idle/idle_in_xact) - node_load_poller: 2Hz /proc/loadavg + /proc/stat procs_running per node Dashboard renderer (sim/render_report.ts + graph.ts): - Util group: one panel per node with translucent orange oversaturation area BEHIND solid blue CPU-util area, 100% reference line, phase-boundary verticals. cols:2 grid wraps after 2 panels per row. - PG node tinted with [PG] flag in legend across the dashboard. - Phase-boundary verticals + push-window shaded zones layered consistently. - All x-axes switched from wall-clock HH:MM to relative seconds-from-bench- start. Shared origin sourced from meta.json's bench_start_ms so 0s on every panel = the same wall-clock moment (previously each chart picked its own earliest sample as origin, causing drift between panels). Oversaturation metric, with explicit fallback: - Primary: (procs_running - ncpu) / ncpu × 100 — true CPU run-queue pressure. - Fallback to load1 when procs_running is missing (older reports). - load1 overcounted previously because it includes uninterruptible D-state procs (PG backends in disk I/O, cgroup_mutex waits), inflating "saturation" by 5-10x under load. - Pure helper extracted to sim/util_metrics.ts; 8 unit tests cover the procs_running > load1 preference, the clamp-at-zero, invalid-ncpu cases. Sampler reliability: - HostPath log file in addition to stdout so the bench's scp-based collector bypasses kubelet log rotation entirely. - main.ts truncates the host log file on every node before pushers start (parallel ssh, best-effort) so it doesn't grow unbounded across runs. - Collector falls back to kubectl-logs when scp fails for any node. Workloads (workloads/): - io_4phase: four-phase IO step (idle → 2.5s → 500ms → 150ms jobs) - io_150ms_flood / io_300ms_flood / io_1s_flood / io_2s_flood: single-phase flood configs to isolate the worker-host CFS context-switch storm vs PG contention regime - burst, ops_day, cpu_*, etc. for other scenarios Tests: - sim/util_metrics_test.ts — 8 cases for computeOversatPct - sim/util_panel_snapshot_test.ts — 5 assertions guarding util-panel SVG invariants (orange behind blue, 100% ref line, relative-time ticks NOT wall-clock, phase-boundary verticals, shared-origin override) Helm values: - sim/values/smoke.yaml — bench-tuned: workers w/ no CPU limit & low mem request, PG w/ 3-core request + wm-critical priorityClass + oomImmune + maxConnections, app w/ wm-critical + oomImmune + no resource limits. - sim/values/local.example.yaml — template for the gitignored local.yaml that carries the EE license key. - Depends on the wm-critical PriorityClass + oomImmune + maxConnections knobs landing in windmill-helm-charts (separate PR). graph.ts additions: - areaFills param: ordered list of per-kind translucent area fills drawn before lines, used by the util panel for orange-behind-blue layering - lineColorOverrides: pin per-kind line colors so oversaturation reliably renders orange regardless of d3 ordinal-color insertion order - highlightKindToken: substring-match flag for the PG-node tint in Node CPU - xRelativeOriginMs: shared bench-start origin for the relative-time x-axis - DataPointMulti is now exported for downstream tests Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
88 lines
3.9 KiB
TypeScript
88 lines
3.9 KiB
TypeScript
// kubectl-based per-second pod timeline. Replaces the cgroup-derived
|
|
// workers-per-node series from the sampler (which had two known bugs: stale
|
|
// pods.json captured only at end of bench, and the sampler's 5s rescan
|
|
// interval missing short-lived pods). Polling kubectl every 1s gives us the
|
|
// authoritative pod inventory at each tick — slower resolution but accurate.
|
|
//
|
|
// Output: JSONL, one line per poll, each line:
|
|
// {"ts": <ms>, "pods": [{"name":"...","node":"...","phase":"..."}, ...]}
|
|
//
|
|
// Stopping: caller flips the `cont` ref to false; the loop exits within ~1s.
|
|
|
|
import { MinikubeProvisioner } from "./k8s_provisioner.ts";
|
|
|
|
export type PodTimelinePoller = {
|
|
// Mutate this to false to stop the loop. The polling promise resolves when
|
|
// the in-flight kubectl call finishes after the flip.
|
|
cont: { value: boolean };
|
|
done: Promise<void>;
|
|
};
|
|
|
|
export function startPodTimeline(
|
|
prov: MinikubeProvisioner,
|
|
outPath: string,
|
|
opts: { intervalMs?: number } = {},
|
|
): PodTimelinePoller {
|
|
const intervalMs = opts.intervalMs ?? 1000;
|
|
const cont = { value: true };
|
|
const f = Deno.openSync(outPath, { write: true, create: true, truncate: true });
|
|
const enc = new TextEncoder();
|
|
|
|
const done = (async () => {
|
|
while (cont.value) {
|
|
const startMs = Date.now();
|
|
try {
|
|
// Capture name|node|phase|ready per pod. `ready` is the first
|
|
// container's readiness gate — that's what kubelet uses for Service
|
|
// routing decisions, and it's the truthy "this worker is actually
|
|
// serving traffic" signal. Excluding the cgroup-still-around-but-
|
|
// CrashLooping case that was making the old chart misleading.
|
|
// CRITICAL: the newline delimiter MUST be the jsonpath string form
|
|
// `{"\n"}` — not literal `\n`. Bare \n means "two-char string '\n'"
|
|
// when emitted by kubectl, so JS split("\n") collapses everything to
|
|
// one mangled row. AND in TypeScript source, the inside-string `\n`
|
|
// must be written as `\\n` so the compiled runtime string preserves
|
|
// the literal `\n` for kubectl to parse — passing a TS template with
|
|
// raw `\n` would put a real newline INSIDE the quoted string,
|
|
// breaking jsonpath's quoted-string parser. Use `\\n` here.
|
|
const res = await prov.kubectl([
|
|
"get", "pods", "-A",
|
|
"-o", "jsonpath={range .items[*]}{.metadata.name}|{.spec.nodeName}|{.status.phase}|{.status.containerStatuses[0].ready}|{.status.containerStatuses[0].restartCount}{\"\\n\"}{end}",
|
|
]);
|
|
if (res.code === 0) {
|
|
const pods: Array<{ name: string; node: string; phase: string; ready: boolean; restarts: number }> = [];
|
|
for (const line of res.stdout.split("\n")) {
|
|
if (!line.trim()) continue;
|
|
const [name, node, phase, ready, restartCount] = line.split("|");
|
|
if (name && node) {
|
|
pods.push({
|
|
name,
|
|
node,
|
|
phase: phase ?? "",
|
|
ready: ready === "true",
|
|
// restartCount lets the renderer detect "pod was restarted
|
|
// mid-bench" events (catches preemption, OOMKill, liveness
|
|
// probe failure, helm rollouts — anything that increments
|
|
// the kubelet's restart counter on the container).
|
|
restarts: Number.isFinite(parseInt(restartCount)) ? parseInt(restartCount) : 0,
|
|
});
|
|
}
|
|
}
|
|
const row = JSON.stringify({ ts: startMs, pods });
|
|
f.writeSync(enc.encode(row + "\n"));
|
|
}
|
|
} catch (_e) {
|
|
// Transient kubectl errors are tolerated — bench should not die because
|
|
// one poll failed. Next iteration retries.
|
|
}
|
|
const elapsed = Date.now() - startMs;
|
|
if (cont.value && elapsed < intervalMs) {
|
|
await new Promise((r) => setTimeout(r, intervalMs - elapsed));
|
|
}
|
|
}
|
|
f.close();
|
|
})();
|
|
|
|
return { cont, done };
|
|
}
|