autocannon-ci can store the results and generate the flamegraphs of your HTTP/1.1 benchmarks of Node.js server. Run your autocannon benchmarks as part of your CI/dev flow, for Node.js.

It can also generate a little website containing all the result of your benchmarking, including flamegraphs with 0x:

• Dashboard
• Job page

npm i autocannon-ci -g

autocannon-ci -c autocannon.yml

server: ./server.js
benchmarks:
  root:
    connections: 100
    duration: 5
    url: localhost:3000
  b:
    connections: 100
    duration: 5
    url: localhost:3000/b
storage:
  type: fs
  path: perf-results

autocannon-ci is a tool to run multiple HTTP/1.1 benchmarks, and generate the relative flamegraphs, with the help of 0x.

Available commands:

• run (default)
• compare
• help

Usage: autocannon-ci run [OPTS]

Runs the benchmarks configured in the autocannon-ci configuration file, and
save them according to the storage configured in the config file. The job id
is used to identify the single run.

Options:

  --config/-c CONFIG      Use the given config file; default: `autocannon.yml`.
  --job/-j ID             Use the specific job id.
  --flamegrah/-F          Generate and store flamegraphs.

Usage: autocannon-ci compare [OPTS] [A] [B]

Compare the job with id A against the job id B. A and B are defaulted to the
latest two jobs.

Options:

  --config/-c CONFIG      Use the given config file; default: `autocannon.yml`.
Launch 'autocannon-ci help [command]' to know more about the commands.

autocannon-ci can store the results and flamegraphs within a storage, which is configured in the config file.

storage:
  type: fs
  path: perf-results

storage:
  type: s3
  bucket: autocannon-ci-test
  region: us-west-2

This will also require the environment variables S3_ACCESS_KEY and S3_SECRET_KEY containing the proper credentials to access S3. It uses the aws-sdk, so any other way of configuring the credential for that will work for autocannon-ci as well.

This project was kindly sponsored by nearForm.

MIT