@nearform/clinic-heap-profilerdeprecated

Programmable interface to Clinic.js Heap Profiler

Usage no npm install needed!

<script type="module">
  import nearformClinicHeapProfiler from 'https://cdn.skypack.dev/@nearform/clinic-heap-profiler';
</script>

README

Clinic.js Heap Profiler

npm version Stability Stable Github Actions build status Downloads Code style

Programmable interface to Clinic.js Heap Profiler. Learn more about Clinic.js: https://clinicjs.org/

Installation

npm i -S @nearform/clinic-heap-profiler

Supported node versions

  • Node.js 10 and above

Example

const ClinicHeapProfiler = require('@nearform/clinic-heap-profiler')
const heapProfiler = new ClinicHeapProfiler()

heapProfiler.collect(['node', './path-to-script.js'], function (err, filepath) {
  if (err) throw err

  heapProfiler.visualize(filepath, filepath + '.html', function (err) {
    if (err) throw err
  })
})

Documentation

const ClinicHeapProfiler = require('@nearform/clinic-heap-profiler')
const heapProfiler = new ClinicHeapProfiler()

new ClinicHeapProfiler([settings])

  • settings <Object>
    • detectPort <boolean> Default: false
    • collectOnFailure <boolean> If set to true, the collected data will be returned even if the process exits with non-zero code. Default: false
    • debug <boolean> If set to true, the generated html will not be minified. Default: false
    • dest <String> The file where the collected data is stored. Default: ./.clinic/<process.pid>.clinic-heapprofile

heapProfiler.collect(args, callback)

Starts a process by using @nearform/heap-profiler.

The process sampling is started as soon as the process starts. The filepath with collected data will be the value in the callback.

stdout, stderr, and stdin will be relayed to the calling process.

The sampling is stopped and data collected right before the process exits.

If you want to collect data earlier, you can send the process a SIGINT or, if detectPort is true, you can call heapProfiler.stopViaIPC().

heapProfiler.visualize(dataFilename, outputFilename, callback)

Will consume the datafile specified by dataFilename, this datafile will be produced by the sampler using heapProfiler.collect.

heapProfiler.visualize will then output a standalone HTML file to outputFilename. When completed the callback will be called with no extra arguments, except a possible error.

heapProfiler.stopViaIPC()

When the profiler is started with detectPort=true, the profiler establish a TCP based IPC communication.

This method can therefore be called to collect the data at any time.

If no TCP channel is opened or available, the method will perform no operation so it is safe to call at all times.

Examples

See the examples folder. All example should be run from the repository main folder:

node examples/redis-web-service

Each index.js will contain any specific setup step required by the example, if any.

License

MIT