Open & closed models
Fixed workers for closed-loop tests, or target a request rate and let concurrency grow with --max-workers to avoid coordinated omission.
Beta · Apache-2.0 · written in Go
resonate drives configurable load against HTTP, HTTP/3, WebSocket, Server-Sent Events, gRPC and MCP servers. One binary for quick one-liners, reusable YAML scenarios, or embed the same engine as a Go library.
resonate hit http://localhost:8080/api/orders -H 'Accept: application/json' --expect-status 200 --rate 50 --duration 30sEverything you need to go from "does it work?" to "does it hold up?" without a pile of tooling.
Fixed workers for closed-loop tests, or target a request rate and let concurrency grow with --max-workers to avoid coordinated omission.
URLs, headers, queries and bodies accept {{ }} expressions re-rendered per request: sequence numbers, UUIDs, env vars, feeders.
Chain requests, extract values from JSON, YAML or XML responses, and reuse them downstream. Define it all in plain YAML.
--assert 'latency_p95<=500ms' returns a non-zero exit code. resonate compare flags regressions between two reports.
A self-contained report.html, a machine-readable report.json, and per-request results.jsonl on every run.
Run several workloads at once, each with its own load profile, plus shared suite defaults and per-step metrics.
Drive requests from files, inline rows or a Kafka topic. Send JSON, protobuf or compressed (gzip, deflate, br, zstd) bodies.
Load-test the tool servers AI agents call. One MCP session per virtual user, every tool call its own step, and isError results counted as failures.
HTTP/1.1, HTTP/2 (including h2c), HTTP/3 over QUIC, mTLS, private CAs, and keep-alive control.
Stop juggling a different load tool for each transport.
Quick runs use resonate hit. When a test deserves to live in version control, describe it in YAML and run it anywhere with resonate run.
resonate schema for editor autocomplete# POST with a templated body, gate on p95 in CI resonate hit https://api.example.com/orders \ -X POST \ -H 'Content-Type: application/json' \ -d '{"id": {{.Seq}}, "tok": "{{uuid}}"}' \ --rate 200 --workers 50 --duration 1m \ --expect-status 200,201 \ --assert 'success_rate>=0.99' \ --assert 'latency_p95<=500ms'
Every run writes a self-contained report.html and a machine-readable report.json. This one is unedited output from a 30s, 60 iterations/s checkout flow (browse → cart → checkout). Excerpt below; open the full report to see the rest.
cat > checkout.yaml << 'EOF'
protocol: http
load:
duration: 30s
rate: 60
workers: 40
http:
timeout: 5s
base_url: http://localhost:18080
flow:
- name: browse
method: GET
url: /products
- name: add-to-cart
method: GET
url: /cart
- name: checkout
method: GET
url: /checkout
EOF
resonate run checkout.yaml
Install with Homebrew, grab a prebuilt binary, or build from source.
brew install jecklgamis/tap/resonateOr build from source with go build -o bin/resonate ./cmd/resonate, or download a binary from the releases page.
resonate hit https://example.com --duration 2sA stdout summary plus report.html, report.json and results.jsonl.
go get github.com/jecklgamis/resonateSame engine, fluent Scenario DSL.
Heads up: resonate is in beta. The scenario YAML schema, report formats and exit codes are stable and versioned; new flags and features may still land, and the Go library API may change before 1.0.
For enquiries, contact us at [email protected]
Pick a starting point, tweak it, and copy the command. Every flag it adds is explained underneath.
resonate hit http://localhost:8080/api/orders \
--rate 50 \
--workers 20 \
--max-workers 200 \
--duration 30s \
--assert 'latency_p95<=500ms'Copy-paste walkthroughs, each against a small throwaway server on your laptop, so you can watch a technique work before pointing it at your own service. No cloud, no accounts, and every one ends with a result you can check.
You are about to ship a change and want CI to block it if performance regresses. Did your last change make the API slower?
cat > api.py << 'EOF'
import os, time
from http.server import ThreadingHTTPServer, BaseHTTPRequestHandler
DELAY = float(os.environ.get("DELAY_MS", "10")) / 1000
class Server(ThreadingHTTPServer):
request_queue_size = 128
class H(BaseHTTPRequestHandler):
protocol_version = "HTTP/1.1"
def do_GET(self):
time.sleep(DELAY) # stands in for your real handler's work
self.send_response(200); self.send_header("Content-Length", "2")
self.end_headers(); self.wfile.write(b"ok")
def log_message(self, *a): pass
Server(("127.0.0.1", 18081), H).serve_forever()
EOFDELAY_MS=10 python3 api.pymkdir baseline
(cd baseline && resonate hit http://127.0.0.1:18081/ --rate 100 --duration 10s)Requests [total, rate] 1098, 109.79/sec Duration [total] 10.001s Success [ratio] 100.00% Latencies [min, mean, p50, p75, p90, p95, p99, max] 10.22ms, 12.465ms, 12.82ms, 12.943ms, 13.115ms, 14.828ms, 15.802ms, 17.628ms Bytes [in, out] 2196, 0 Status Codes [code:count] 200:1098
DELAY_MS=14 python3 api.pymkdir candidate
(cd candidate && resonate hit http://127.0.0.1:18081/ --rate 100 --duration 10s)Requests [total, rate] 1099, 109.90/sec Duration [total] 10s Success [ratio] 99.91% Latencies [min, mean, p50, p75, p90, p95, p99, max] 14.346ms, 17.583ms, 17.859ms, 17.94ms, 18.039ms, 23.495ms, 25.362ms, 26.648ms Bytes [in, out] 2196, 0 Status Codes [code:count] 200:1098 Errors: [1] Get "http://127.0.0.1:18081/": context deadline exceeded
resonate compare baseline/report.json candidate/report.json --fail-on 'latency_p95>20%'
echo "exit code: $?"Metric Baseline Current Change requests 1098 1099 +0.1% rate 109.79 109.90 +0.1% success_rate 100.00% 99.91% -0.1% error_rate 0.00% 0.09% n/a latency_mean 12.465ms 17.583ms +41.1% latency_p50 12.82ms 17.859ms +39.3% latency_p75 12.943ms 17.94ms +38.6% latency_p90 13.115ms 18.039ms +37.5% latency_p95 14.828ms 23.495ms +58.5% latency_p99 15.802ms 25.362ms +60.5% FAIL latency_p95>20%: 14.828ms -> 23.495ms (+58.5%) error: 1 of 1 regression rule(s) matched exit code: 1
DELAY_MS=10 python3 api.pymkdir rollback
(cd rollback && resonate hit http://127.0.0.1:18081/ --rate 100 --duration 10s)
resonate compare baseline/report.json rollback/report.json --fail-on 'latency_p95>20%'
echo "exit code: $?"Metric Baseline Current Change requests 1098 1098 +0.0% rate 109.79 109.79 +0.0% success_rate 100.00% 100.00% +0.0% error_rate 0.00% 0.00% +0.0% latency_mean 12.465ms 12.544ms +0.6% latency_p50 12.82ms 12.861ms +0.3% latency_p75 12.943ms 12.976ms +0.3% latency_p90 13.115ms 13.14ms +0.2% latency_p95 14.828ms 15.917ms +7.3% latency_p99 15.802ms 16.876ms +6.8% ok latency_p95>20%: 14.828ms -> 15.917ms (+7.3%) exit code: 0
What you'll see. Every run writes a report.json, and compare turns two of them into a verdict: the rule latency_p95>20% means fail when p95 is more than 20% slower than the baseline. Here the 4 ms slowdown is +58% and fails with exit code 1; the rollback passes. Pick a threshold above your noise: an unchanged rerun of this very test drifted about 7% on p95, so a 10% rule would have flaked. In CI, keep the baseline report.json from your main branch, run compare in the pull-request job, and a non-zero exit fails the build. You can gate one step too, for example login:latency_p99>20ms.