Languages: [English] [简体中文]

This project is based on egg-prometheus, with more performance metrics. It provides more Prometheus functionalities for egg. For more information, please refer to Node.js monitering solutions.

$ npm i egg-exporter --save

Start the plugin with configurations in ${app_root}/config/plugin.js:

exports.exporter = {
  enable: true,
  package: 'egg-exporter',
};

exports.exporter = {
  scrapePort: 3000,
  scrapePath: '/metrics',
  prefix: 'egg_',
  defaultLabels: { stage: 'dev' },
};

• scrapePort: the port used to scrape metrics
• scrapePath: the path used for monitoring metrics
• prefix: the prefix for the metrics names
• defaultLabels: default metrics labels, globally effective
• aggregatorPort: the TCP port for RPC

• http_request_duration_milliseconds histogram: http request duration
• http_request_size_bytes summary: http request body size
• http_response_size_bytes summary: http reponse body size
• http_request_total counter: number of http requests
• http_all_errors_total counter: number of http erorrs
• http_all_request_in_processing_total gauge: number of http requests being processed
• process_resident_memory_bytes gauge: resident memory size
• nodejs_heap_size_total_bytes gauge: allocated heap memory size
• nodejs_heap_size_used_bytes gauge: allocated stack memory size
• nodejs_external_memory_bytes gauge: C++ bind objects memory usage
• nodejs_version_info: version information

When the egg-rpc-base plugin is enabled, the following metrics are also provided

• rpc_consumer_response_time_ms summary: rpc client response time in milliseconds
• rpc_consumer_request_rate counter: rpc client number of requests
• rpc_consumer_fail_response_time_ms summary: rpc client failed response time
• rpc_consumer_request_fail_rate counter: rpc failed requests
• rpc_consumer_request_size_bytes summary: rpc request size
• rpc_consumer_response_size_bytes summary: rpc response size
• rpc_provider_response_time_ms summary: rpc server response time
• rpc_provider_request_rate counter: rpc server number of requests received
• rpc_provider_fail_response_time_ms summary: rpc server failed response time in milliseconds
• rpc_provider_request_fail_rate counter: rpc server failed responses

You may use the following API to customize metrics in your business logics.

const counter = new app.prometheus.Counter({
  name: 'xxx_total',
  help: 'custom counter',
  labelNames: [ 'xxx' ],
});

const gauge = new app.prometheus.Gauge({
  name: 'xxx_gauge',
  help: 'custom gauge',
  labelNames: [ 'xxx' ],
});

const histogram = new app.prometheus.Histogram({
  name: 'xxx_histogram',
  help: 'custom histogram',
  labelNames: [ 'xxx' ],
});

const summary = new app.prometheus.Summary({
  name: 'xxx_summary',
  help: 'custom summary',
  labelNames: [ 'xxx' ],
});

Please tell us what we can do for you, but before that, check if there are existing bugs or issues.

MIT