Cheatwrk HTTP load tester
Light · Postcard
102 × 152 mm · PDF
Dark · Postcard
102 × 152 mm · PDF
Ink saver · Postcard
102 × 152 mm · PDF
Light · A4 landscape
297 × 210 mm · PDF
Dark · A4 landscape
297 × 210 mm · PDF
Ink saver · A4 landscape
297 × 210 mm · PDF
Light · A3 landscape
420 × 297 mm · PDF
Dark · A3 landscape
420 × 297 mm · PDF
Ink saver · A3 landscape
420 × 297 mm · PDF
Start here
wrk -t2 -c20 -d10s <url>
Quick 10 s test
wrk -t12 -c400 -d30s <url>
Heavier load
wrk --latency <url>
Percentile latencies
wrk -H "<header>" <url>
Add a header
wrk -s <script.lua> <url>
Run a Lua script
wrk --timeout 5s <url>
Count slow replies
wrk -c<N> -t<N> -d<time>
The three core knobs
wrk -s <script.lua> <url> -- <args>
Pass script args
Concepts
- setup once per thread
- init per thread, gets args
- request / response every call
- done final summary
Running a test
wrk -t<n> -c<n> -d<time>
<url>
Run a benchmark
-t, --threads <n>
Worker threads
-c, --connections <n>
Open connections in total
-d, --duration <time>
Test length (2s, 2m, 2h)
--latencyPrint latency distribution
--timeout <time>
Timeout per response
-H, --header "<name: value>"
Add a request header
-s, --script <file>
Load a Lua script
-- <arg>...
Pass args to init(args)
-v, --versionShow version
Reading output
| Line | Meaning |
|---|---|
| Latency | Per-request time: avg, stdev, max |
| Req/Sec | Per-thread rate, not total |
| Requests/sec | Total throughput |
| Transfer/sec | Bytes read per second |
| Socket errors | connect/read/write/timeout counts |
| Non-2xx or 3xx | Responses with error status |
Lua scripting
wrk.method = "POST"
Set the HTTP method
wrk.body = '<data>'
Set the request body
wrk.headers["<name>"] =
"<value>"
Set a header
wrk.path / wrk.host / wrk.port
Target parts (read)
function request()
Return the next request
wrk.format(method, path, headers, body)
Build a request string
function response(status, headers, body)
Inspect each response
function init(args)
Per-thread setup
function setup(thread)
Run once per thread
function done(summary, latency, requests)
Custom final report
function delay()
Return ms to wait
latency:percentile(99.0)
Latency percentile in done()
summary.errors.timeout
Error counters in done()
thread:set("<k>", v) /
thread:get("<k>")
Share data with threads
Recipes
wrk -t4 -c100 -d30s --latency http://localhost:8080/wrk.method = "POST"wrk.body = '{"name":"x"}'wrk.headers["Content-Type"] = "application/json"# wrk -t2 -c20 -d10s -s post.lua http://localhost:8080/itemscounter = 0request = function() counter = counter + 1 return wrk.format("GET", "/item/" .. counter)endresponse = function(status, headers, body) if status ~= 200 then print(status) endend
