3.2 KiB
k6 load test
k6/sse.js holds 120 concurrent Server-Sent Events connections for 5 minutes
and reports connection count plus p95 event latency.
Why a custom k6 binary
Native k6 buffers a complete HTTP response, so it cannot parse a stream that
never ends. The script imports the community k6/x/sse extension
(github.com/phymbert/xk6-sse). Newer k6 releases resolve community extensions
automatically. If the stock grafana/k6 image rejects the import, build a
binary with the extension:
make k6-build # creates k6/k6-sse (git-ignored)
K6=./k6/k6-sse make load
make load uses, in order: k6 on PATH, k6/k6-sse, then
docker run --rm -i grafana/k6.
Procedure
-
Confirm the API is healthy and the SSE route works:
curl -sS https://api.pulse.noelsabia.dev/api/healthz curl -sS -N --max-time 20 https://api.pulse.noelsabia.dev/api/streamYou should see
retry: 3000and ascenarioevent immediately, then apingevery 15 s. -
Run the load test:
make load # or: API_URL=https://api.pulse.noelsabia.dev k6 run k6/sse.jsmake loadwritesk6/results/summary-<timestamp>.json(git-ignored). -
During the run, post a scenario change so the scenario-event latency has samples and the propagation path is exercised:
curl -u "$CONTROL_USER:$CONTROL_PASS" -X POST \ -H 'Content-Type: application/json' \ -d '{"scenario":"sunny-saturday-market","weather":{"temp_delta_c":4,"precip_mm":0}}' \ https://api.pulse.noelsabia.dev/api/scenarioRepeat every 30–60 s with different scenarios if you want a distribution. Without a change,
sse_event_latencyis measured from heartbeatpingevents, which still proves the stream is live. -
Record these numbers in the paper/README from the saved summary:
sse_connections_openedcount (must be 120)sse_connections_activemax (must reach 120)sse_open_successrate (must be >= 99%)sse_event_latencyp95 (target < 2000 ms)sse_connection_errors(target 0)sse_events_receivedtotal
Pass criteria
| Metric | Target |
|---|---|
| Connections opened | 120 |
| Peak active connections | 120 |
| Open success rate | >= 99% |
| p95 event latency | < 2000 ms |
| Unexpected connection errors | 0 |
| Duration | 5 minutes |
The k6 thresholds enforce the first, third, and fourth; the others are read
from the summary. A p95 above 2 s means Cloudflare, the Go SSE hub, or the
client is buffering/delaying events; check X-Accel-Buffering: no, the 15 s
heartbeat, and whether the connection is HTTP/2-buffered at the edge.
Notes
sse_event_latencyuses the server timestamp indata.t(ping) ordata.at(scenario). It includes clock skew between the k6 host and the pod; keep both on NTP. Negative or > 1 h values are ignored.- Each VU closes after 20 heartbeats (20 × 15 s = 5 min), so the run ends cleanly instead of waiting for the 7-minute request timeout.
- One API replica means all clients share one SSE hub; a pod restart drops
every stream and clients reconnect after
retry: 3000. - k6 itself does not collect server CPU/memory. Use
kubectl top pods(if metrics-server is available) or watch the pod during the run.