This repo is intended to allow comparison of the performance and simplicity (maintenance costs) of various API backend languages and frameworks.
Setup modules and get the fiber module:
cd go-fiber
go mod init github.com/bernhard-hofmann/ApiSmackdown
go get github.com/gofiber/fiber
go run main.gocd node-express
npm init
npm install express --saveWas created with
$HOME/dotnet/dotnet new web -o dotnet-6Run it with
$HOME/dotnet/dotnet runFollow the setup on the k6 website.
To save the reader time, I'll publish results here. Note, these are from my laptop and are, as such, completely non-scientific. They may offer the reader some insight because the tests will be run one after the other at the same time on the same hardware.
bernhard@Latitude-5591:~/github/bernhard-hofmann/ApiSmackdown/k6$ k6 run go-fiber.js
/\ |‾‾| /‾‾/ /‾‾/
/\ / \ | |/ / / /
/ \/ \ | ( / ‾‾\
/ \ | |\ \ | (‾) |
/ __________ \ |__| \__\ \_____/ .io
execution: local
script: go-fiber.js
output: -
scenarios: (100.00%) 1 scenario, 10 max VUs, 1m0s max duration (incl. graceful stop):
* default: 10 looping VUs for 30s (gracefulStop: 30s)
running (0m30.0s), 00/10 VUs, 2329701 complete and 0 interrupted iterations
default ✓ [======================================] 10 VUs 30s
data_received..................: 331 MB 11 MB/s
data_sent......................: 186 MB 6.2 MB/s
http_req_blocked...............: avg=1.51µs min=646ns med=1.34µs max=2.74ms p(90)=1.74µs p(95)=2.19µs
http_req_connecting............: avg=0ns min=0s med=0s max=234.17µs p(90)=0s p(95)=0s
http_req_duration..............: avg=82.88µs min=27.63µs med=73.66µs max=9.31ms p(90)=113.01µs p(95)=131.31µs
{ expected_response:true }...: avg=82.88µs min=27.63µs med=73.66µs max=9.31ms p(90)=113.01µs p(95)=131.31µs
http_req_failed................: 0.00% ✓ 0 ✗ 2329701
http_req_receiving.............: avg=16.5µs min=5.36µs med=14.69µs max=7.76ms p(90)=18.63µs p(95)=23.29µs
http_req_sending...............: avg=6.69µs min=2.83µs med=6.2µs max=5.75ms p(90)=7.58µs p(95)=9.2µs
http_req_tls_handshaking.......: avg=0s min=0s med=0s max=0s p(90)=0s p(95)=0s
http_req_waiting...............: avg=59.67µs min=15.92µs med=51.95µs max=9.26ms p(90)=88.42µs p(95)=101.33µs
http_reqs......................: 2329701 77654.472525/s
iteration_duration.............: avg=123.6µs min=47.97µs med=107.48µs max=426ms p(90)=151.25µs p(95)=181.39µs
iterations.....................: 2329701 77654.472525/s
vus............................: 10 min=10 max=10
vus_max........................: 10 min=10 max=10
bernhard@Latitude-5591:~/github/bernhard-hofmann/ApiSmackdown/k6$ k6 run node-express.js
/\ |‾‾| /‾‾/ /‾‾/
/\ / \ | |/ / / /
/ \/ \ | ( / ‾‾\
/ \ | |\ \ | (‾) |
/ __________ \ |__| \__\ \_____/ .io
execution: local
script: node-express.js
output: -
scenarios: (100.00%) 1 scenario, 10 max VUs, 1m0s max duration (incl. graceful stop):
* default: 10 looping VUs for 30s (gracefulStop: 30s)
running (0m30.0s), 00/10 VUs, 621677 complete and 0 interrupted iterations
default ✓ [======================================] 10 VUs 30s
data_received..................: 143 MB 4.8 MB/s
data_sent......................: 50 MB 1.7 MB/s
http_req_blocked...............: avg=951ns min=618ns med=763ns max=493.02µs p(90)=1.33µs p(95)=1.65µs
http_req_connecting............: avg=0ns min=0s med=0s max=84.06µs p(90)=0s p(95)=0s
http_req_duration..............: avg=455.68µs min=79.53µs med=382.11µs max=5.91ms p(90)=748.61µs p(95)=813.53µs
{ expected_response:true }...: avg=455.68µs min=79.53µs med=382.11µs max=5.91ms p(90)=748.61µs p(95)=813.53µs
http_req_failed................: 0.00% ✓ 0 ✗ 621677
http_req_receiving.............: avg=12.59µs min=6.02µs med=10.42µs max=3.98ms p(90)=16.86µs p(95)=20.11µs
http_req_sending...............: avg=4.42µs min=2.93µs med=3.54µs max=1.7ms p(90)=5.84µs p(95)=6.93µs
http_req_tls_handshaking.......: avg=0s min=0s med=0s max=0s p(90)=0s p(95)=0s
http_req_waiting...............: avg=438.66µs min=52.68µs med=366.53µs max=4.06ms p(90)=732.7µs p(95)=794.53µs
http_reqs......................: 621677 20721.795817/s
iteration_duration.............: avg=479.62µs min=115.99µs med=404.48µs max=5.94ms p(90)=771.18µs p(95)=839.84µs
iterations.....................: 621677 20721.795817/s
vus............................: 10 min=10 max=10
vus_max........................: 10 min=10 max=10
bernhard@Latitude-5591:~/github/bernhard-hofmann/ApiSmackdown/k6$ k6 run dotnet-6.js
/\ |‾‾| /‾‾/ /‾‾/
/\ / \ | |/ / / /
/ \/ \ | ( / ‾‾\
/ \ | |\ \ | (‾) |
/ __________ \ |__| \__\ \_____/ .io
execution: local
script: dotnet-6.js
output: -
scenarios: (100.00%) 1 scenario, 10 max VUs, 1m0s max duration (incl. graceful stop):
* default: 10 looping VUs for 30s (gracefulStop: 30s)
running (0m30.0s), 00/10 VUs, 1653449 complete and 0 interrupted iterations
default ✓ [======================================] 10 VUs 30s
data_received..................: 293 MB 9.8 MB/s
data_sent......................: 132 MB 4.4 MB/s
http_req_blocked...............: avg=1.83µs min=572ns med=1.39µs max=6.85ms p(90)=1.8µs p(95)=2.43µs
http_req_connecting............: avg=0ns min=0s med=0s max=93.52µs p(90)=0s p(95)=0s
http_req_duration..............: avg=131.28µs min=46.58µs med=100.58µs max=44.75ms p(90)=182.64µs p(95)=244.4µs
{ expected_response:true }...: avg=131.28µs min=46.58µs med=100.58µs max=44.75ms p(90)=182.64µs p(95)=244.4µs
http_req_failed................: 0.00% ✓ 0 ✗ 1653449
http_req_receiving.............: avg=23.93µs min=6.89µs med=16.32µs max=15.77ms p(90)=22.94µs p(95)=27.93µs
http_req_sending...............: avg=8.16µs min=2.97µs med=6.39µs max=14.09ms p(90)=8.07µs p(95)=11.42µs
http_req_tls_handshaking.......: avg=0s min=0s med=0s max=0s p(90)=0s p(95)=0s
http_req_waiting...............: avg=99.18µs min=30.45µs med=74.95µs max=44.65ms p(90)=149.28µs p(95)=199.14µs
http_reqs......................: 1653449 55113.291313/s
iteration_duration.............: avg=175.74µs min=68.23µs med=138.04µs max=44.99ms p(90)=231.55µs p(95)=309µs
iterations.....................: 1653449 55113.291313/s
vus............................: 10 min=10 max=10
vus_max........................: 10 min=10 max=10
