Compare commits
13 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
| ed05c01abc | |||
| 8fd8b9a915 | |||
| 56757e1c71 | |||
| d2ee396cf7 | |||
| 5a22b87592 | |||
| b99d2cc04b | |||
| 32d75804c0 | |||
| 7c3197455f | |||
| 767a40e01e | |||
| fc017d31b0 | |||
| f67b093127 | |||
| 6bcdf73a97 | |||
| c47732c5fc |
40
changelog.md
40
changelog.md
@@ -1,5 +1,45 @@
|
||||
# Changelog
|
||||
|
||||
## 2026-03-02 - 3.0.2 - fix(pidusage)
|
||||
prune history entries for PIDs not present in the requested set to avoid stale data and memory growth
|
||||
|
||||
- Deletes entries from the history map when a PID is not included in the current pids array
|
||||
- Prevents accumulation of stale PID histories and potential memory growth
|
||||
- Change implemented in ts/smartmetrics.pidusage.ts alongside the metrics result construction
|
||||
|
||||
## 2026-02-19 - 3.0.1 - fix(smartmetrics)
|
||||
no code changes detected; no version bump or release required
|
||||
|
||||
- git diff contained no modifications
|
||||
- current package.json version is 3.0.0
|
||||
- no dependency or file changes to warrant a release
|
||||
|
||||
## 2026-02-19 - 3.0.0 - BREAKING CHANGE(smartmetrics)
|
||||
add system-wide metrics collection, Prometheus gauges, and normalized CPU reporting
|
||||
|
||||
- Add new sysusage plugin (ts/smartmetrics.sysusage.ts) that reads /proc/stat and /proc/meminfo (with os fallback) and returns system CPU, memory and load averages.
|
||||
- Expose system-wide Prometheus gauges: smartmetrics_system_cpu_percent, smartmetrics_system_memory_used_percent, smartmetrics_system_memory_used_bytes, smartmetrics_system_load_avg_1, smartmetrics_system_load_avg_5, smartmetrics_system_load_avg_15.
|
||||
- Extend IMetricsSnapshot with system fields: systemCpuPercent, systemMemTotalBytes, systemMemAvailableBytes, systemMemUsedBytes, systemMemUsedPercent, systemLoadAvg1, systemLoadAvg5, systemLoadAvg15 (this is a breaking TypeScript API change).
|
||||
- Normalize per-process CPU in pidusage by adding cpuCoreCount and cpuNormalizedPercent and use cpuNormalizedPercent when aggregating CPU across the process tree.
|
||||
- Export the new sysusage plugin from ts/smartmetrics.plugins.ts and wire system metrics into metric collection and Prometheus gauge updates.
|
||||
|
||||
## 2026-02-19 - 2.0.11 - fix(deps)
|
||||
bump dependencies, update build script, expand README and npm metadata
|
||||
|
||||
- Bumped runtime deps: @push.rocks/smartdelay ^3.0.5, @push.rocks/smartlog ^3.1.11
|
||||
- Updated devDependencies: @git.zone/tsbuild, tsbundle, tsrun, tstest and @types/node versions
|
||||
- Changed build script: "(tsbuild --web)" → "(tsbuild tsfolders)"
|
||||
- Updated npmextra.json: renamed keys (gitzone → @git.zone/cli, tsdoc → @git.zone/tsdoc), added release registries and accessLevel, and added @ship.zone/szci entry
|
||||
- Extensive README improvements: installation notes (pnpm), clearer API docs, examples, added Issue Reporting & Security section and utility docs (formatBytes)
|
||||
|
||||
## 2025-06-09 - 2.0.9 - fix(readme)
|
||||
Update documentation with detailed usage instructions, API references and integration examples.
|
||||
|
||||
- Overhauled README to provide a clearer explanation of SmartMetrics features and API.
|
||||
- Added a quick start guide, detailed examples, and code snippets for various integrations (Express, PM2, custom dashboards).
|
||||
- Reorganized documentation sections to better highlight core concepts including process aggregation and memory limit detection.
|
||||
- Updated installation instructions and usage examples to reflect the latest functionality.
|
||||
|
||||
## 2025-06-09 - 2.0.8 - fix(smartmetrics)
|
||||
Refactor metrics calculation and update Prometheus integration documentation
|
||||
|
||||
|
||||
@@ -1,5 +1,5 @@
|
||||
{
|
||||
"gitzone": {
|
||||
"@git.zone/cli": {
|
||||
"projectType": "npm",
|
||||
"module": {
|
||||
"githost": "code.foss.global",
|
||||
@@ -25,13 +25,19 @@
|
||||
"log reporting",
|
||||
"operational insights"
|
||||
]
|
||||
},
|
||||
"release": {
|
||||
"registries": [
|
||||
"https://verdaccio.lossless.digital",
|
||||
"https://registry.npmjs.org"
|
||||
],
|
||||
"accessLevel": "public"
|
||||
}
|
||||
},
|
||||
"npmci": {
|
||||
"npmGlobalTools": [],
|
||||
"npmAccessLevel": "public"
|
||||
},
|
||||
"tsdoc": {
|
||||
"@git.zone/tsdoc": {
|
||||
"legal": "\n## License and Legal Information\n\nThis repository contains open-source code that is licensed under the MIT License. A copy of the MIT License can be found in the [license](license) file within this repository. \n\n**Please note:** The MIT License does not grant permission to use the trade names, trademarks, service marks, or product names of the project, except as required for reasonable and customary use in describing the origin of the work and reproducing the content of the NOTICE file.\n\n### Trademarks\n\nThis project is owned and maintained by Task Venture Capital GmbH. The names and logos associated with Task Venture Capital GmbH and any related products or services are trademarks of Task Venture Capital GmbH and are not included within the scope of the MIT license granted herein. Use of these trademarks must comply with Task Venture Capital GmbH's Trademark Guidelines, and any usage must be approved in writing by Task Venture Capital GmbH.\n\n### Company Information\n\nTask Venture Capital GmbH \nRegistered at District court Bremen HRB 35230 HB, Germany\n\nFor any legal inquiries or if you require further information, please contact us via email at hello@task.vc.\n\nBy using this repository, you acknowledge that you have read this section, agree to comply with its terms, and understand that the licensing of the code does not imply endorsement by Task Venture Capital GmbH of any derivative works.\n"
|
||||
},
|
||||
"@ship.zone/szci": {
|
||||
"npmGlobalTools": []
|
||||
}
|
||||
}
|
||||
22
package.json
22
package.json
@@ -1,6 +1,6 @@
|
||||
{
|
||||
"name": "@push.rocks/smartmetrics",
|
||||
"version": "2.0.8",
|
||||
"version": "3.0.2",
|
||||
"private": false,
|
||||
"description": "A package for easy collection and reporting of system and process metrics.",
|
||||
"main": "dist_ts/index.js",
|
||||
@@ -9,15 +9,15 @@
|
||||
"license": "MIT",
|
||||
"scripts": {
|
||||
"test": "(tstest test/ --verbose)",
|
||||
"build": "(tsbuild --web)",
|
||||
"build": "(tsbuild tsfolders)",
|
||||
"buildDocs": "tsdoc"
|
||||
},
|
||||
"devDependencies": {
|
||||
"@git.zone/tsbuild": "^2.6.4",
|
||||
"@git.zone/tsbundle": "^2.0.8",
|
||||
"@git.zone/tsrun": "^1.2.44",
|
||||
"@git.zone/tstest": "^2.3.1",
|
||||
"@types/node": "^22.15.30"
|
||||
"@git.zone/tsbuild": "^4.1.2",
|
||||
"@git.zone/tsbundle": "^2.8.3",
|
||||
"@git.zone/tsrun": "^2.0.1",
|
||||
"@git.zone/tstest": "^3.1.8",
|
||||
"@types/node": "^25.3.0"
|
||||
},
|
||||
"browserslist": [
|
||||
"last 1 chrome versions"
|
||||
@@ -35,12 +35,8 @@
|
||||
"readme.md"
|
||||
],
|
||||
"dependencies": {
|
||||
"@push.rocks/smartdelay": "^3.0.1",
|
||||
"@push.rocks/smartlog": "^3.0.2",
|
||||
"@types/pidusage": "^2.0.2",
|
||||
"pidtree": "^0.6.0",
|
||||
"pidusage": "^4.0.1",
|
||||
"prom-client": "^15.1.3"
|
||||
"@push.rocks/smartdelay": "^3.0.5",
|
||||
"@push.rocks/smartlog": "^3.1.11"
|
||||
},
|
||||
"type": "module",
|
||||
"keywords": [
|
||||
|
||||
4722
pnpm-lock.yaml
generated
4722
pnpm-lock.yaml
generated
File diff suppressed because it is too large
Load Diff
324
readme.md
324
readme.md
@@ -1,147 +1,285 @@
|
||||
# @push.rocks/smartmetrics
|
||||
|
||||
easy system metrics
|
||||
|
||||
## Install
|
||||
**Powerful system metrics collection for Node.js applications with Prometheus integration** 🚀
|
||||
|
||||
To include `@push.rocks/smartmetrics` in your project, you need a Node.js environment with support for ES Modules. Make sure that your `package.json` contains `"type": "module"`. You can install `@push.rocks/smartmetrics` using npm:
|
||||
## Issue Reporting and Security
|
||||
|
||||
For reporting bugs, issues, or security vulnerabilities, please visit [community.foss.global/](https://community.foss.global/). This is the central community hub for all issue reporting. Developers who sign and comply with our contribution agreement and go through identification can also get a [code.foss.global/](https://code.foss.global/) account to submit Pull Requests directly.
|
||||
|
||||
## What is SmartMetrics?
|
||||
|
||||
SmartMetrics is a comprehensive metrics collection library that monitors your Node.js application's resource usage in real-time. It tracks CPU usage, memory consumption, and system metrics across your main process and all child processes, providing insights through both JSON and Prometheus formats.
|
||||
|
||||
## Key Features
|
||||
|
||||
- 📊 **Real-time Metrics Collection** – Monitor CPU and memory usage across all processes
|
||||
- 🔄 **Automatic Child Process Tracking** – Aggregates metrics from main and child processes via `pidtree` + `pidusage`
|
||||
- 🐳 **Docker-Aware** – Detects container memory limits from cgroup automatically
|
||||
- 📈 **Prometheus Integration** – Built-in HTTP endpoint for Prometheus scraping with `prom-client`
|
||||
- 🔧 **Flexible Output Formats** – Get metrics as JSON objects or Prometheus text exposition format
|
||||
- 📝 **Automatic Heartbeat Logging** – Optional periodic metrics logging via `@push.rocks/smartlog`
|
||||
- 🚀 **Zero Configuration** – Works out of the box with sensible defaults
|
||||
|
||||
## Installation
|
||||
|
||||
```bash
|
||||
npm install @push.rocks/smartmetrics --save
|
||||
pnpm install @push.rocks/smartmetrics
|
||||
# or
|
||||
npm install @push.rocks/smartmetrics
|
||||
```
|
||||
|
||||
Or using yarn:
|
||||
|
||||
```bash
|
||||
yarn add @push.rocks/smartmetrics
|
||||
```
|
||||
|
||||
## Usage
|
||||
|
||||
The `@push.rocks/smartmetrics` package provides an easy way to collect and monitor system metrics such as CPU and memory usage within your Node.js applications. It's built with TypeScript, offering type safety and IntelliSense in editors that support it.
|
||||
|
||||
### Getting Started
|
||||
|
||||
Before diving into the code, ensure you have `@push.rocks/smartmetrics` installed in your project and your environment is configured to use ES Modules (ESM) and TypeScript.
|
||||
|
||||
#### Initialization
|
||||
|
||||
To begin using `smartmetrics`, you need to import the module and initialize it:
|
||||
## Quick Start
|
||||
|
||||
```typescript
|
||||
// Import dependencies
|
||||
import { SmartMetrics } from '@push.rocks/smartmetrics';
|
||||
import { Smartlog } from '@push.rocks/smartlog';
|
||||
|
||||
// Setup a logger (or use your existing logger if applicable)
|
||||
// Create a logger instance
|
||||
const logger = new Smartlog({
|
||||
logContext: 'applicationName',
|
||||
logContext: null,
|
||||
minimumLogLevel: 'info',
|
||||
});
|
||||
logger.enableConsole();
|
||||
|
||||
// Initialize smartmetrics with the prepared logger
|
||||
const smartMetrics = new SmartMetrics(logger, 'sourceName');
|
||||
// Initialize SmartMetrics
|
||||
const metrics = new SmartMetrics(logger, 'my-service');
|
||||
|
||||
// Get metrics on-demand
|
||||
const currentMetrics = await metrics.getMetrics();
|
||||
console.log(`CPU Usage: ${currentMetrics.cpuUsageText}`);
|
||||
console.log(`Memory: ${currentMetrics.memoryUsageText}`);
|
||||
|
||||
// Enable automatic heartbeat logging (every 20 seconds)
|
||||
metrics.start();
|
||||
|
||||
// Enable Prometheus endpoint
|
||||
metrics.enablePrometheusEndpoint(9090);
|
||||
// Metrics now available at http://localhost:9090/metrics
|
||||
|
||||
// Clean shutdown
|
||||
metrics.stop();
|
||||
```
|
||||
|
||||
In this snippet, `Smartlog` is used for logging purposes. Replace `'applicationName'` and `'sourceName'` with appropriate values for your application.
|
||||
## Core Concepts
|
||||
|
||||
#### Starting Metrics Monitoring
|
||||
### Process Aggregation
|
||||
|
||||
With `smartMetrics` initialized, you can start the metrics monitoring process:
|
||||
SmartMetrics doesn't just monitor your main process – it automatically discovers and aggregates metrics from all child processes spawned by your application using `pidtree`. This gives you a complete picture of your application's resource footprint, not just the parent process.
|
||||
|
||||
### Memory Limit Detection
|
||||
|
||||
The library automatically detects available memory whether running on bare metal, in Docker containers, or with Node.js heap restrictions. It picks the most restrictive of:
|
||||
|
||||
1. **System total memory** (`os.totalmem()`)
|
||||
2. **Docker cgroup limit** – supports both cgroup v2 (`/sys/fs/cgroup/memory.max`) and cgroup v1 (`/sys/fs/cgroup/memory/memory.limit_in_bytes`)
|
||||
3. **V8 heap size limit** (`v8.getHeapStatistics().heap_size_limit`)
|
||||
|
||||
This ensures accurate percentage calculations regardless of environment.
|
||||
|
||||
### Dual Output Formats
|
||||
|
||||
- **JSON Format** (`getMetrics()`) – Ideal for application monitoring, custom dashboards, and programmatic access
|
||||
- **Prometheus Format** (`getPrometheusFormattedMetrics()`) – Perfect for integration with Prometheus/Grafana monitoring stacks
|
||||
|
||||
## API Reference
|
||||
|
||||
### `new SmartMetrics(logger, sourceName)`
|
||||
|
||||
Creates a new SmartMetrics instance.
|
||||
|
||||
| Parameter | Type | Description |
|
||||
|-----------|------|-------------|
|
||||
| `logger` | `Smartlog` | A `@push.rocks/smartlog` logger instance |
|
||||
| `sourceName` | `string` | Identifier for your service/application |
|
||||
|
||||
### `async getMetrics(): Promise<IMetricsSnapshot>`
|
||||
|
||||
Retrieves current system metrics as a structured object.
|
||||
|
||||
**Returns `IMetricsSnapshot`:**
|
||||
|
||||
```typescript
|
||||
{
|
||||
process_cpu_seconds_total: number; // Total CPU time in seconds
|
||||
nodejs_active_handles_total: number; // Always 0 (deprecated Node.js API; real values tracked by Prometheus default collectors)
|
||||
nodejs_active_requests_total: number; // Always 0 (deprecated Node.js API; real values tracked by Prometheus default collectors)
|
||||
nodejs_heap_size_total_bytes: number; // V8 heap size in bytes
|
||||
cpuPercentage: number; // Aggregated CPU usage across all child processes
|
||||
cpuUsageText: string; // Human-readable CPU usage (e.g. "12.5 %")
|
||||
memoryPercentage: number; // Memory usage as percentage of detected limit
|
||||
memoryUsageBytes: number; // Total memory in bytes across all child processes
|
||||
memoryUsageText: string; // Human-readable (e.g. "45% | 920 MB / 2 GB")
|
||||
}
|
||||
```
|
||||
|
||||
**Example:**
|
||||
|
||||
```typescript
|
||||
const metrics = await smartMetrics.getMetrics();
|
||||
if (metrics.cpuPercentage > 80) {
|
||||
console.warn('High CPU usage detected!');
|
||||
}
|
||||
```
|
||||
|
||||
### `start(): void`
|
||||
|
||||
Starts automatic metrics collection and heartbeat logging every 20 seconds via the provided `Smartlog` instance.
|
||||
|
||||
```typescript
|
||||
smartMetrics.start();
|
||||
// Logs: "sending heartbeat for my-service with metrics" every 20 seconds
|
||||
```
|
||||
|
||||
This starts the collection and monitoring of various system metrics. By default, `smartMetrics` sends heartbeat messages including these metrics at a regular interval.
|
||||
### `stop(): void`
|
||||
|
||||
#### Metrics Collection
|
||||
Stops automatic metrics collection, closes heartbeat loop, and shuts down any open Prometheus endpoints.
|
||||
|
||||
To manually collect metrics at any point, you can call the `getMetrics` method. This could be useful for logging or sending metrics to a monitoring tool:
|
||||
### `async getPrometheusFormattedMetrics(): Promise<string>`
|
||||
|
||||
Returns all metrics in Prometheus text exposition format, including default Node.js collectors and custom SmartMetrics gauges.
|
||||
|
||||
```typescript
|
||||
async function logMetrics() {
|
||||
const metrics = await smartMetrics.getMetrics();
|
||||
console.log(metrics);
|
||||
}
|
||||
|
||||
logMetrics();
|
||||
const promMetrics = await smartMetrics.getPrometheusFormattedMetrics();
|
||||
// Returns:
|
||||
// # HELP smartmetrics_cpu_percentage Current CPU usage percentage
|
||||
// # TYPE smartmetrics_cpu_percentage gauge
|
||||
// smartmetrics_cpu_percentage 15.2
|
||||
// # HELP smartmetrics_memory_percentage Current memory usage percentage
|
||||
// # TYPE smartmetrics_memory_percentage gauge
|
||||
// smartmetrics_memory_percentage 45.3
|
||||
// # HELP smartmetrics_memory_usage_bytes Current memory usage in bytes
|
||||
// # TYPE smartmetrics_memory_usage_bytes gauge
|
||||
// smartmetrics_memory_usage_bytes 965214208
|
||||
// ... plus all default Node.js metrics from prom-client
|
||||
```
|
||||
|
||||
#### Customizing Monitoring
|
||||
### `enablePrometheusEndpoint(port?: number): void`
|
||||
|
||||
The `start` method begins an unattended process of collecting and logging metrics. If you require more control over how and when metrics are collected or reported, you can use the `getMetrics` method in combination with your logic.
|
||||
Starts an HTTP server that exposes metrics at `/metrics` for Prometheus scraping.
|
||||
|
||||
#### Stopping Metrics Monitoring
|
||||
|
||||
To stop the automatic metrics monitoring, simply call:
|
||||
| Parameter | Type | Default | Description |
|
||||
|-----------|------|---------|-------------|
|
||||
| `port` | `number` | `9090` | Port to listen on |
|
||||
|
||||
```typescript
|
||||
smartMetrics.stop();
|
||||
```
|
||||
|
||||
This stops the internal loop that periodically collects and logs metrics. It's useful for clean shutdowns of your application or when metrics monitoring is only needed during specific intervals.
|
||||
|
||||
### Understanding the Metrics
|
||||
|
||||
The `getMetrics` method returns a snapshot of various system metrics, including CPU and memory usage. Here's a brief overview of the information provided:
|
||||
|
||||
- `process_cpu_seconds_total`: Total CPU time spent by the process.
|
||||
- `nodejs_active_handles_total`: Number of active handles.
|
||||
- `nodejs_active_requests_total`: Number of active requests.
|
||||
- `nodejs_heap_size_total_bytes`: Total size of the heap.
|
||||
- `cpuPercentage`: Overall CPU usage percentage.
|
||||
- `cpuUsageText`: Readable string representation of CPU usage.
|
||||
- `memoryPercentage`: Percentage of memory used.
|
||||
- `memoryUsageBytes`: Total memory used in bytes.
|
||||
- `memoryUsageText`: Readable string representation of memory usage.
|
||||
|
||||
### Prometheus Integration
|
||||
|
||||
`smartmetrics` can expose metrics in Prometheus format for scraping:
|
||||
|
||||
```typescript
|
||||
// Enable Prometheus endpoint on port 9090 (default)
|
||||
smartMetrics.enablePrometheusEndpoint();
|
||||
|
||||
// Or specify a custom port
|
||||
smartMetrics.enablePrometheusEndpoint(3000);
|
||||
|
||||
// The metrics will be available at http://localhost:9090/metrics
|
||||
|
||||
// To get Prometheus-formatted metrics programmatically
|
||||
const prometheusMetrics = await smartMetrics.getPrometheusFormattedMetrics();
|
||||
console.log(prometheusMetrics);
|
||||
|
||||
// Disable the endpoint when done
|
||||
smartMetrics.disablePrometheusEndpoint();
|
||||
// GET http://localhost:3000/metrics → Prometheus text format
|
||||
// GET http://localhost:3000/anything-else → 404
|
||||
```
|
||||
|
||||
The Prometheus endpoint exposes both default Node.js metrics (via prom-client) and custom calculated metrics:
|
||||
- `smartmetrics_cpu_percentage` - Current CPU usage percentage
|
||||
- `smartmetrics_memory_percentage` - Current memory usage percentage
|
||||
- `smartmetrics_memory_usage_bytes` - Current memory usage in bytes
|
||||
- Plus all default Node.js metrics collected by prom-client
|
||||
### `disablePrometheusEndpoint(): void`
|
||||
|
||||
### Conclusion
|
||||
Gracefully shuts down the Prometheus HTTP server.
|
||||
|
||||
`@push.rocks/smartmetrics` offers a straightforward and efficient way to monitor essential system metrics of your Node.js application. By integrating it, you gain valuable insights into the performance and health of your system, aiding in diagnosis and optimization efforts.
|
||||
### `formatBytes(bytes, decimals?): string`
|
||||
|
||||
Utility method to convert byte values to human-readable strings.
|
||||
|
||||
```typescript
|
||||
smartMetrics.formatBytes(1073741824); // "1 GB"
|
||||
smartMetrics.formatBytes(1536, 1); // "1.5 KB"
|
||||
```
|
||||
|
||||
## Use Cases
|
||||
|
||||
### 🔍 Application Performance Monitoring
|
||||
|
||||
```typescript
|
||||
const metricsBefore = await smartMetrics.getMetrics();
|
||||
await performHeavyOperation();
|
||||
const metricsAfter = await smartMetrics.getMetrics();
|
||||
|
||||
console.log(`Operation consumed ${
|
||||
metricsAfter.process_cpu_seconds_total - metricsBefore.process_cpu_seconds_total
|
||||
} CPU seconds`);
|
||||
```
|
||||
|
||||
### 🛡️ Resource Limit Enforcement
|
||||
|
||||
```typescript
|
||||
async function checkResources() {
|
||||
const metrics = await smartMetrics.getMetrics();
|
||||
|
||||
if (metrics.memoryPercentage > 90) {
|
||||
throw new Error('Memory usage too high, refusing new operations');
|
||||
}
|
||||
|
||||
if (metrics.cpuPercentage > 95) {
|
||||
await delay(1000); // Back off when CPU is stressed
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### 📈 Prometheus + Grafana Stack
|
||||
|
||||
```typescript
|
||||
smartMetrics.enablePrometheusEndpoint(9090);
|
||||
|
||||
// prometheus.yml:
|
||||
// scrape_configs:
|
||||
// - job_name: 'my-app'
|
||||
// scrape_interval: 15s
|
||||
// static_configs:
|
||||
// - targets: ['localhost:9090']
|
||||
```
|
||||
|
||||
### 🐳 Container Resource Monitoring
|
||||
|
||||
```typescript
|
||||
// Automatically detects Docker/cgroup memory limits
|
||||
const metrics = await smartMetrics.getMetrics();
|
||||
console.log(metrics.memoryUsageText);
|
||||
// Output: "45% | 920 MB / 2 GB" (container limit detected)
|
||||
```
|
||||
|
||||
### 🔄 Health Check Endpoint
|
||||
|
||||
```typescript
|
||||
import express from 'express';
|
||||
|
||||
const app = express();
|
||||
|
||||
app.get('/health', async (req, res) => {
|
||||
const metrics = await smartMetrics.getMetrics();
|
||||
|
||||
res.json({
|
||||
status: metrics.memoryPercentage < 90 ? 'healthy' : 'degraded',
|
||||
cpu: metrics.cpuUsageText,
|
||||
memory: metrics.memoryUsageText,
|
||||
});
|
||||
});
|
||||
```
|
||||
|
||||
### 🔁 Graceful Restart on High Memory (PM2)
|
||||
|
||||
```typescript
|
||||
setInterval(async () => {
|
||||
const metrics = await smartMetrics.getMetrics();
|
||||
|
||||
if (metrics.memoryPercentage > 95) {
|
||||
console.error('Memory limit reached, requesting restart');
|
||||
process.exit(0); // PM2 will restart the process
|
||||
}
|
||||
}, 10000);
|
||||
```
|
||||
|
||||
## License and Legal Information
|
||||
|
||||
This repository contains open-source code that is licensed under the MIT License. A copy of the MIT License can be found in the [license](license) file within this repository.
|
||||
This repository contains open-source code licensed under the MIT License. A copy of the license can be found in the [LICENSE](./LICENSE) file.
|
||||
|
||||
**Please note:** The MIT License does not grant permission to use the trade names, trademarks, service marks, or product names of the project, except as required for reasonable and customary use in describing the origin of the work and reproducing the content of the NOTICE file.
|
||||
|
||||
### Trademarks
|
||||
|
||||
This project is owned and maintained by Task Venture Capital GmbH. The names and logos associated with Task Venture Capital GmbH and any related products or services are trademarks of Task Venture Capital GmbH and are not included within the scope of the MIT license granted herein. Use of these trademarks must comply with Task Venture Capital GmbH's Trademark Guidelines, and any usage must be approved in writing by Task Venture Capital GmbH.
|
||||
This project is owned and maintained by Task Venture Capital GmbH. The names and logos associated with Task Venture Capital GmbH and any related products or services are trademarks of Task Venture Capital GmbH or third parties, and are not included within the scope of the MIT license granted herein.
|
||||
|
||||
Use of these trademarks must comply with Task Venture Capital GmbH's Trademark Guidelines or the guidelines of the respective third-party owners, and any usage must be approved in writing. Third-party trademarks used herein are the property of their respective owners and used only in a descriptive manner, e.g. for an implementation of an API or similar.
|
||||
|
||||
### Company Information
|
||||
|
||||
Task Venture Capital GmbH
|
||||
Registered at District court Bremen HRB 35230 HB, Germany
|
||||
Task Venture Capital GmbH
|
||||
Registered at District Court Bremen HRB 35230 HB, Germany
|
||||
|
||||
For any legal inquiries or if you require further information, please contact us via email at hello@task.vc.
|
||||
For any legal inquiries or further information, please contact us via email at hello@task.vc.
|
||||
|
||||
By using this repository, you acknowledge that you have read this section, agree to comply with its terms, and understand that the licensing of the code does not imply endorsement by Task Venture Capital GmbH of any derivative works.
|
||||
|
||||
@@ -75,4 +75,4 @@ tap.test('should disable Prometheus endpoint', async () => {
|
||||
}
|
||||
});
|
||||
|
||||
tap.start();
|
||||
export default tap.start();
|
||||
|
||||
@@ -3,6 +3,6 @@
|
||||
*/
|
||||
export const commitinfo = {
|
||||
name: '@push.rocks/smartmetrics',
|
||||
version: '2.0.8',
|
||||
version: '3.0.2',
|
||||
description: 'A package for easy collection and reporting of system and process metrics.'
|
||||
}
|
||||
|
||||
@@ -5,41 +5,82 @@ export class SmartMetrics {
|
||||
public started = false;
|
||||
public sourceNameArg: string;
|
||||
public logger: plugins.smartlog.Smartlog;
|
||||
public registry: plugins.promClient.Registry;
|
||||
public registry: plugins.prom.Registry;
|
||||
public maxMemoryMB: number;
|
||||
|
||||
|
||||
// Prometheus gauges for custom metrics
|
||||
private cpuPercentageGauge: plugins.promClient.Gauge<string>;
|
||||
private memoryPercentageGauge: plugins.promClient.Gauge<string>;
|
||||
private memoryUsageBytesGauge: plugins.promClient.Gauge<string>;
|
||||
|
||||
private cpuPercentageGauge: plugins.prom.Gauge;
|
||||
private memoryPercentageGauge: plugins.prom.Gauge;
|
||||
private memoryUsageBytesGauge: plugins.prom.Gauge;
|
||||
private systemCpuPercentGauge: plugins.prom.Gauge;
|
||||
private systemMemUsedPercentGauge: plugins.prom.Gauge;
|
||||
private systemMemUsedBytesGauge: plugins.prom.Gauge;
|
||||
private systemLoadAvg1Gauge: plugins.prom.Gauge;
|
||||
private systemLoadAvg5Gauge: plugins.prom.Gauge;
|
||||
private systemLoadAvg15Gauge: plugins.prom.Gauge;
|
||||
|
||||
// HTTP server for Prometheus endpoint
|
||||
private prometheusServer?: plugins.http.Server;
|
||||
private prometheusPort?: number;
|
||||
|
||||
public async setup() {
|
||||
const collectDefaultMetrics = plugins.promClient.collectDefaultMetrics;
|
||||
this.registry = new plugins.promClient.Registry();
|
||||
collectDefaultMetrics({ register: this.registry });
|
||||
|
||||
public setup() {
|
||||
this.registry = new plugins.prom.Registry();
|
||||
plugins.prom.collectDefaultMetrics(this.registry);
|
||||
|
||||
// Initialize custom gauges
|
||||
this.cpuPercentageGauge = new plugins.promClient.Gauge({
|
||||
this.cpuPercentageGauge = new plugins.prom.Gauge({
|
||||
name: 'smartmetrics_cpu_percentage',
|
||||
help: 'Current CPU usage percentage',
|
||||
registers: [this.registry]
|
||||
});
|
||||
|
||||
this.memoryPercentageGauge = new plugins.promClient.Gauge({
|
||||
|
||||
this.memoryPercentageGauge = new plugins.prom.Gauge({
|
||||
name: 'smartmetrics_memory_percentage',
|
||||
help: 'Current memory usage percentage',
|
||||
registers: [this.registry]
|
||||
});
|
||||
|
||||
this.memoryUsageBytesGauge = new plugins.promClient.Gauge({
|
||||
|
||||
this.memoryUsageBytesGauge = new plugins.prom.Gauge({
|
||||
name: 'smartmetrics_memory_usage_bytes',
|
||||
help: 'Current memory usage in bytes',
|
||||
registers: [this.registry]
|
||||
});
|
||||
|
||||
this.systemCpuPercentGauge = new plugins.prom.Gauge({
|
||||
name: 'smartmetrics_system_cpu_percent',
|
||||
help: 'System-wide CPU usage percentage',
|
||||
registers: [this.registry]
|
||||
});
|
||||
|
||||
this.systemMemUsedPercentGauge = new plugins.prom.Gauge({
|
||||
name: 'smartmetrics_system_memory_used_percent',
|
||||
help: 'System-wide memory usage percentage',
|
||||
registers: [this.registry]
|
||||
});
|
||||
|
||||
this.systemMemUsedBytesGauge = new plugins.prom.Gauge({
|
||||
name: 'smartmetrics_system_memory_used_bytes',
|
||||
help: 'System-wide memory used in bytes',
|
||||
registers: [this.registry]
|
||||
});
|
||||
|
||||
this.systemLoadAvg1Gauge = new plugins.prom.Gauge({
|
||||
name: 'smartmetrics_system_load_avg_1',
|
||||
help: 'System 1-minute load average',
|
||||
registers: [this.registry]
|
||||
});
|
||||
|
||||
this.systemLoadAvg5Gauge = new plugins.prom.Gauge({
|
||||
name: 'smartmetrics_system_load_avg_5',
|
||||
help: 'System 5-minute load average',
|
||||
registers: [this.registry]
|
||||
});
|
||||
|
||||
this.systemLoadAvg15Gauge = new plugins.prom.Gauge({
|
||||
name: 'smartmetrics_system_load_avg_15',
|
||||
help: 'System 15-minute load average',
|
||||
registers: [this.registry]
|
||||
});
|
||||
}
|
||||
|
||||
constructor(loggerArg: plugins.smartlog.Smartlog, sourceNameArg: string) {
|
||||
@@ -50,28 +91,32 @@ export class SmartMetrics {
|
||||
}
|
||||
|
||||
private checkMemoryLimits() {
|
||||
let heapStats = plugins.v8.getHeapStatistics();
|
||||
let maxHeapSizeMB = heapStats.heap_size_limit / 1024 / 1024;
|
||||
let totalSystemMemoryMB = plugins.os.totalmem() / 1024 / 1024;
|
||||
const heapStats = plugins.v8.getHeapStatistics();
|
||||
const maxHeapSizeMB = heapStats.heap_size_limit / 1024 / 1024;
|
||||
const totalSystemMemoryMB = plugins.os.totalmem() / 1024 / 1024;
|
||||
|
||||
let dockerMemoryLimitMB = totalSystemMemoryMB;
|
||||
|
||||
// Try cgroup v2 first, then fall back to cgroup v1
|
||||
try {
|
||||
let dockerMemoryLimitBytes = plugins.fs.readFileSync(
|
||||
'/sys/fs/cgroup/memory/memory.limit_in_bytes',
|
||||
'utf8'
|
||||
);
|
||||
dockerMemoryLimitMB = parseInt(dockerMemoryLimitBytes, 10) / 1024 / 1024;
|
||||
} catch (error) {
|
||||
// Ignore - this will fail if not running in a Docker container
|
||||
const cgroupV2 = plugins.fs.readFileSync('/sys/fs/cgroup/memory.max', 'utf8').trim();
|
||||
if (cgroupV2 !== 'max') {
|
||||
dockerMemoryLimitMB = parseInt(cgroupV2, 10) / 1024 / 1024;
|
||||
}
|
||||
} catch {
|
||||
try {
|
||||
const cgroupV1 = plugins.fs.readFileSync(
|
||||
'/sys/fs/cgroup/memory/memory.limit_in_bytes',
|
||||
'utf8'
|
||||
).trim();
|
||||
dockerMemoryLimitMB = parseInt(cgroupV1, 10) / 1024 / 1024;
|
||||
} catch {
|
||||
// Not running in a container — use system memory
|
||||
}
|
||||
}
|
||||
|
||||
// Set the maximum memory to the lower value between the Docker limit and the total system memory
|
||||
// Pick the most restrictive limit
|
||||
this.maxMemoryMB = Math.min(totalSystemMemoryMB, dockerMemoryLimitMB, maxHeapSizeMB);
|
||||
|
||||
// If the maximum old space size limit is greater than the maximum available memory, throw an error
|
||||
if (maxHeapSizeMB > this.maxMemoryMB) {
|
||||
throw new Error('Node.js process can use more memory than is available');
|
||||
}
|
||||
}
|
||||
|
||||
public start() {
|
||||
@@ -104,20 +149,26 @@ export class SmartMetrics {
|
||||
}
|
||||
|
||||
public async getMetrics() {
|
||||
const pids = await plugins.pidtree(process.pid);
|
||||
const stats = await plugins.pidusage([process.pid, ...pids]);
|
||||
let pids: number[] = [];
|
||||
try {
|
||||
pids = await plugins.pidtree.getChildPids(process.pid);
|
||||
} catch {
|
||||
// pidtree can fail if process tree cannot be read
|
||||
}
|
||||
const stats = await plugins.pidusage.getPidUsage([process.pid, ...pids]);
|
||||
|
||||
// Aggregate normalized CPU (0-100% of total machine) across process tree
|
||||
let cpuPercentage = 0;
|
||||
for (const stat of Object.keys(stats)) {
|
||||
if (!stats[stat]) continue;
|
||||
cpuPercentage += stats[stat].cpu;
|
||||
for (const stat of Object.values(stats)) {
|
||||
if (!stat) continue;
|
||||
cpuPercentage += stat.cpuNormalizedPercent;
|
||||
}
|
||||
let cpuUsageText = `${Math.round(cpuPercentage * 100) / 100} %`;
|
||||
|
||||
let memoryUsageBytes = 0;
|
||||
for (const stat of Object.keys(stats)) {
|
||||
if (!stats[stat]) continue;
|
||||
memoryUsageBytes += stats[stat].memory;
|
||||
for (const stat of Object.values(stats)) {
|
||||
if (!stat) continue;
|
||||
memoryUsageBytes += stat.memory;
|
||||
}
|
||||
|
||||
let memoryPercentage =
|
||||
@@ -126,8 +177,9 @@ export class SmartMetrics {
|
||||
memoryUsageBytes
|
||||
)} / ${this.formatBytes(this.maxMemoryMB * 1024 * 1024)}`;
|
||||
|
||||
console.log(`${cpuUsageText} ||| ${memoryUsageText} `);
|
||||
|
||||
// Get system-wide metrics
|
||||
const systemUsage = await plugins.sysusage.getSystemUsage();
|
||||
|
||||
// Update Prometheus gauges with current values
|
||||
if (this.cpuPercentageGauge) {
|
||||
this.cpuPercentageGauge.set(cpuPercentage);
|
||||
@@ -138,19 +190,35 @@ export class SmartMetrics {
|
||||
if (this.memoryUsageBytesGauge) {
|
||||
this.memoryUsageBytesGauge.set(memoryUsageBytes);
|
||||
}
|
||||
if (this.systemCpuPercentGauge) {
|
||||
this.systemCpuPercentGauge.set(systemUsage.cpuPercent);
|
||||
}
|
||||
if (this.systemMemUsedPercentGauge) {
|
||||
this.systemMemUsedPercentGauge.set(systemUsage.memUsedPercent);
|
||||
}
|
||||
if (this.systemMemUsedBytesGauge) {
|
||||
this.systemMemUsedBytesGauge.set(systemUsage.memUsedBytes);
|
||||
}
|
||||
if (this.systemLoadAvg1Gauge) {
|
||||
this.systemLoadAvg1Gauge.set(systemUsage.loadAvg1);
|
||||
}
|
||||
if (this.systemLoadAvg5Gauge) {
|
||||
this.systemLoadAvg5Gauge.set(systemUsage.loadAvg5);
|
||||
}
|
||||
if (this.systemLoadAvg15Gauge) {
|
||||
this.systemLoadAvg15Gauge.set(systemUsage.loadAvg15);
|
||||
}
|
||||
|
||||
// Calculate Node.js metrics directly
|
||||
const cpuUsage = process.cpuUsage();
|
||||
const process_cpu_seconds_total = (cpuUsage.user + cpuUsage.system) / 1000000; // Convert from microseconds to seconds
|
||||
|
||||
const process_cpu_seconds_total = (cpuUsage.user + cpuUsage.system) / 1000000;
|
||||
|
||||
const heapStats = plugins.v8.getHeapStatistics();
|
||||
const nodejs_heap_size_total_bytes = heapStats.total_heap_size;
|
||||
|
||||
// Note: Active handles and requests are internal Node.js metrics that require deprecated APIs
|
||||
// We return 0 here, but the Prometheus default collectors will track the real values
|
||||
|
||||
const nodejs_active_handles_total = 0;
|
||||
const nodejs_active_requests_total = 0;
|
||||
|
||||
|
||||
const returnMetrics: interfaces.IMetricsSnapshot = {
|
||||
process_cpu_seconds_total,
|
||||
nodejs_active_handles_total,
|
||||
@@ -161,6 +229,14 @@ export class SmartMetrics {
|
||||
memoryPercentage,
|
||||
memoryUsageBytes,
|
||||
memoryUsageText,
|
||||
systemCpuPercent: systemUsage.cpuPercent,
|
||||
systemMemTotalBytes: systemUsage.memTotalBytes,
|
||||
systemMemAvailableBytes: systemUsage.memAvailableBytes,
|
||||
systemMemUsedBytes: systemUsage.memUsedBytes,
|
||||
systemMemUsedPercent: systemUsage.memUsedPercent,
|
||||
systemLoadAvg1: systemUsage.loadAvg1,
|
||||
systemLoadAvg5: systemUsage.loadAvg5,
|
||||
systemLoadAvg15: systemUsage.loadAvg15,
|
||||
};
|
||||
return returnMetrics;
|
||||
}
|
||||
@@ -168,7 +244,7 @@ export class SmartMetrics {
|
||||
public async getPrometheusFormattedMetrics(): Promise<string> {
|
||||
// Update metrics to ensure gauges have latest values
|
||||
await this.getMetrics();
|
||||
|
||||
|
||||
// Return Prometheus text exposition format
|
||||
return await this.registry.metrics();
|
||||
}
|
||||
@@ -178,7 +254,7 @@ export class SmartMetrics {
|
||||
this.logger.log('warn', 'Prometheus endpoint is already running');
|
||||
return;
|
||||
}
|
||||
|
||||
|
||||
this.prometheusServer = plugins.http.createServer(async (req, res) => {
|
||||
if (req.url === '/metrics' && req.method === 'GET') {
|
||||
try {
|
||||
@@ -195,7 +271,7 @@ export class SmartMetrics {
|
||||
res.end('Not Found');
|
||||
}
|
||||
});
|
||||
|
||||
|
||||
this.prometheusPort = port;
|
||||
this.prometheusServer.listen(port, () => {
|
||||
this.logger.log('info', `Prometheus metrics endpoint available at http://localhost:${port}/metrics`);
|
||||
@@ -206,12 +282,12 @@ export class SmartMetrics {
|
||||
if (!this.prometheusServer) {
|
||||
return;
|
||||
}
|
||||
|
||||
|
||||
const port = this.prometheusPort;
|
||||
this.prometheusServer.close(() => {
|
||||
this.logger.log('info', `Prometheus metrics endpoint on port ${port} has been shut down`);
|
||||
});
|
||||
|
||||
|
||||
this.prometheusServer = undefined;
|
||||
this.prometheusPort = undefined;
|
||||
}
|
||||
|
||||
@@ -1,11 +1,21 @@
|
||||
export interface IMetricsSnapshot {
|
||||
// existing process/node fields
|
||||
process_cpu_seconds_total: number;
|
||||
nodejs_active_handles_total: number;
|
||||
nodejs_active_requests_total: number;
|
||||
nodejs_heap_size_total_bytes: number;
|
||||
cpuPercentage: number;
|
||||
cpuPercentage: number; // normalized to 0-100% of total machine
|
||||
cpuUsageText: string;
|
||||
memoryPercentage: number;
|
||||
memoryUsageBytes: number;
|
||||
memoryUsageText: string;
|
||||
// system-wide fields
|
||||
systemCpuPercent: number;
|
||||
systemMemTotalBytes: number;
|
||||
systemMemAvailableBytes: number;
|
||||
systemMemUsedBytes: number;
|
||||
systemMemUsedPercent: number;
|
||||
systemLoadAvg1: number;
|
||||
systemLoadAvg5: number;
|
||||
systemLoadAvg15: number;
|
||||
}
|
||||
|
||||
55
ts/smartmetrics.pidtree.ts
Normal file
55
ts/smartmetrics.pidtree.ts
Normal file
@@ -0,0 +1,55 @@
|
||||
import * as fs from 'fs';
|
||||
|
||||
// Get all descendant PIDs of the given root PID by reading /proc/<pid>/stat.
|
||||
// Returns an array of descendant PIDs (excludes the root itself).
|
||||
export async function getChildPids(rootPid: number): Promise<number[]> {
|
||||
const parentMap = new Map<number, number[]>(); // parent → children
|
||||
|
||||
let entries: string[];
|
||||
try {
|
||||
entries = fs.readdirSync('/proc');
|
||||
} catch {
|
||||
return [];
|
||||
}
|
||||
|
||||
for (const entry of entries) {
|
||||
const pid = parseInt(entry, 10);
|
||||
if (isNaN(pid)) continue;
|
||||
|
||||
try {
|
||||
const stat = fs.readFileSync(`/proc/${pid}/stat`, 'utf8');
|
||||
// Format: pid (comm) state ppid ...
|
||||
// comm can contain spaces and parentheses, so find the last ')' first
|
||||
const closeParenIdx = stat.lastIndexOf(')');
|
||||
if (closeParenIdx === -1) continue;
|
||||
const afterComm = stat.slice(closeParenIdx + 2); // skip ') '
|
||||
const fields = afterComm.split(' ');
|
||||
const ppid = parseInt(fields[1], 10); // field index 1 after state is ppid
|
||||
|
||||
if (!parentMap.has(ppid)) {
|
||||
parentMap.set(ppid, []);
|
||||
}
|
||||
parentMap.get(ppid)!.push(pid);
|
||||
} catch {
|
||||
// Process may have exited between readdir and readFile
|
||||
continue;
|
||||
}
|
||||
}
|
||||
|
||||
// BFS from rootPid to collect all descendants
|
||||
const result: number[] = [];
|
||||
const queue: number[] = [rootPid];
|
||||
|
||||
while (queue.length > 0) {
|
||||
const current = queue.shift()!;
|
||||
const children = parentMap.get(current);
|
||||
if (children) {
|
||||
for (const child of children) {
|
||||
result.push(child);
|
||||
queue.push(child);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return result;
|
||||
}
|
||||
136
ts/smartmetrics.pidusage.ts
Normal file
136
ts/smartmetrics.pidusage.ts
Normal file
@@ -0,0 +1,136 @@
|
||||
import * as fs from 'fs';
|
||||
import * as os from 'os';
|
||||
import { execSync } from 'child_process';
|
||||
|
||||
// CPU core count (cached at module load)
|
||||
const cpuCoreCount = typeof os.availableParallelism === 'function'
|
||||
? os.availableParallelism()
|
||||
: os.cpus().length;
|
||||
|
||||
// Cached system constants
|
||||
let clkTck: number | null = null;
|
||||
let pageSize: number | null = null;
|
||||
|
||||
function getClkTck(): number {
|
||||
if (clkTck === null) {
|
||||
try {
|
||||
clkTck = parseInt(execSync('getconf CLK_TCK', { encoding: 'utf8' }).trim(), 10);
|
||||
} catch {
|
||||
clkTck = 100; // standard Linux default
|
||||
}
|
||||
}
|
||||
return clkTck;
|
||||
}
|
||||
|
||||
function getPageSize(): number {
|
||||
if (pageSize === null) {
|
||||
try {
|
||||
pageSize = parseInt(execSync('getconf PAGESIZE', { encoding: 'utf8' }).trim(), 10);
|
||||
} catch {
|
||||
pageSize = 4096; // standard Linux default
|
||||
}
|
||||
}
|
||||
return pageSize;
|
||||
}
|
||||
|
||||
// History for CPU delta tracking
|
||||
interface ISnapshot {
|
||||
utime: number;
|
||||
stime: number;
|
||||
timestamp: number; // hrtime in seconds
|
||||
}
|
||||
|
||||
const history = new Map<number, ISnapshot>();
|
||||
|
||||
function readProcStat(pid: number): { utime: number; stime: number; rss: number } | null {
|
||||
try {
|
||||
const stat = fs.readFileSync(`/proc/${pid}/stat`, 'utf8');
|
||||
// Format: pid (comm) state ppid ... fields
|
||||
// utime is field 14, stime is field 15, rss is field 24 (1-indexed)
|
||||
const closeParenIdx = stat.lastIndexOf(')');
|
||||
if (closeParenIdx === -1) return null;
|
||||
const afterComm = stat.slice(closeParenIdx + 2);
|
||||
const fields = afterComm.split(' ');
|
||||
// fields[0] = state (field 3), so utime = fields[11] (field 14), stime = fields[12] (field 15), rss = fields[21] (field 24)
|
||||
const utime = parseInt(fields[11], 10);
|
||||
const stime = parseInt(fields[12], 10);
|
||||
const rss = parseInt(fields[21], 10);
|
||||
return { utime, stime, rss };
|
||||
} catch {
|
||||
return null;
|
||||
}
|
||||
}
|
||||
|
||||
function hrtimeSeconds(): number {
|
||||
const [sec, nsec] = process.hrtime();
|
||||
return sec + nsec / 1e9;
|
||||
}
|
||||
|
||||
export interface IPidUsageResult {
|
||||
cpu: number; // raw per-core CPU% (can exceed 100%)
|
||||
cpuCoreCount: number; // number of CPU cores on the machine
|
||||
cpuNormalizedPercent: number; // cpu / coreCount — 0-100% of total machine
|
||||
memory: number; // RSS in bytes
|
||||
}
|
||||
|
||||
/**
|
||||
* Get CPU percentage and memory usage for the given PIDs.
|
||||
* CPU% is calculated as a delta between successive calls.
|
||||
*/
|
||||
export async function getPidUsage(
|
||||
pids: number[]
|
||||
): Promise<Record<number, IPidUsageResult>> {
|
||||
const tck = getClkTck();
|
||||
const ps = getPageSize();
|
||||
const result: Record<number, IPidUsageResult> = {};
|
||||
|
||||
for (const pid of pids) {
|
||||
const stat = readProcStat(pid);
|
||||
if (!stat) {
|
||||
continue;
|
||||
}
|
||||
|
||||
const now = hrtimeSeconds();
|
||||
const totalTicks = stat.utime + stat.stime;
|
||||
const memoryBytes = stat.rss * ps;
|
||||
|
||||
const prev = history.get(pid);
|
||||
if (prev) {
|
||||
const elapsedSeconds = now - prev.timestamp;
|
||||
const ticksDelta = totalTicks - (prev.utime + prev.stime);
|
||||
const cpuSeconds = ticksDelta / tck;
|
||||
const cpuPercent = elapsedSeconds > 0 ? (cpuSeconds / elapsedSeconds) * 100 : 0;
|
||||
|
||||
result[pid] = {
|
||||
cpu: cpuPercent,
|
||||
cpuCoreCount,
|
||||
cpuNormalizedPercent: cpuPercent / cpuCoreCount,
|
||||
memory: memoryBytes,
|
||||
};
|
||||
} else {
|
||||
// First call for this PID — no delta available, report 0% cpu
|
||||
result[pid] = {
|
||||
cpu: 0,
|
||||
cpuCoreCount,
|
||||
cpuNormalizedPercent: 0,
|
||||
memory: memoryBytes,
|
||||
};
|
||||
}
|
||||
|
||||
// Update history
|
||||
history.set(pid, {
|
||||
utime: stat.utime,
|
||||
stime: stat.stime,
|
||||
timestamp: now,
|
||||
});
|
||||
}
|
||||
|
||||
// Prune history entries for PIDs no longer in the requested set
|
||||
for (const histPid of history.keys()) {
|
||||
if (!pids.includes(histPid)) {
|
||||
history.delete(histPid);
|
||||
}
|
||||
}
|
||||
|
||||
return result;
|
||||
}
|
||||
@@ -12,9 +12,10 @@ import * as smartlog from '@push.rocks/smartlog';
|
||||
|
||||
export { smartdelay, smartlog };
|
||||
|
||||
// third party scope
|
||||
import pidusage from 'pidusage';
|
||||
import pidtree from 'pidtree';
|
||||
import * as promClient from 'prom-client';
|
||||
// own implementations (replacing pidtree, pidusage, prom-client)
|
||||
import * as pidtree from './smartmetrics.pidtree.js';
|
||||
import * as pidusage from './smartmetrics.pidusage.js';
|
||||
import * as prom from './smartmetrics.prom.js';
|
||||
import * as sysusage from './smartmetrics.sysusage.js';
|
||||
|
||||
export { pidusage, pidtree, promClient };
|
||||
export { pidtree, pidusage, prom, sysusage };
|
||||
|
||||
671
ts/smartmetrics.prom.ts
Normal file
671
ts/smartmetrics.prom.ts
Normal file
@@ -0,0 +1,671 @@
|
||||
import * as v8 from 'v8';
|
||||
import * as fs from 'fs';
|
||||
import { PerformanceObserver, monitorEventLoopDelay } from 'perf_hooks';
|
||||
|
||||
// ── Metric types ────────────────────────────────────────────────────────────
|
||||
|
||||
export interface IGaugeConfig {
|
||||
name: string;
|
||||
help: string;
|
||||
registers?: Registry[];
|
||||
labelNames?: string[];
|
||||
collect?: () => void | Promise<void>;
|
||||
}
|
||||
|
||||
export interface ICounterConfig {
|
||||
name: string;
|
||||
help: string;
|
||||
registers?: Registry[];
|
||||
labelNames?: string[];
|
||||
collect?: () => void | Promise<void>;
|
||||
}
|
||||
|
||||
export interface IHistogramConfig {
|
||||
name: string;
|
||||
help: string;
|
||||
registers?: Registry[];
|
||||
labelNames?: string[];
|
||||
buckets?: number[];
|
||||
collect?: () => void | Promise<void>;
|
||||
}
|
||||
|
||||
interface IMetric {
|
||||
name: string;
|
||||
help: string;
|
||||
type: string;
|
||||
collect?: () => void | Promise<void>;
|
||||
getLines(): Promise<string[]>;
|
||||
}
|
||||
|
||||
// ── Registry ────────────────────────────────────────────────────────────────
|
||||
|
||||
export class Registry {
|
||||
private metricsList: IMetric[] = [];
|
||||
|
||||
registerMetric(metric: IMetric): void {
|
||||
this.metricsList.push(metric);
|
||||
}
|
||||
|
||||
async metrics(): Promise<string> {
|
||||
const lines: string[] = [];
|
||||
for (const m of this.metricsList) {
|
||||
if (m.collect) {
|
||||
await m.collect();
|
||||
}
|
||||
lines.push(`# HELP ${m.name} ${m.help}`);
|
||||
lines.push(`# TYPE ${m.name} ${m.type}`);
|
||||
lines.push(...(await m.getLines()));
|
||||
}
|
||||
return lines.join('\n') + '\n';
|
||||
}
|
||||
}
|
||||
|
||||
// ── Gauge ───────────────────────────────────────────────────────────────────
|
||||
|
||||
export class Gauge implements IMetric {
|
||||
public name: string;
|
||||
public help: string;
|
||||
public type = 'gauge';
|
||||
public collect?: () => void | Promise<void>;
|
||||
private value = 0;
|
||||
private labelledValues = new Map<string, number>();
|
||||
constructor(config: IGaugeConfig) {
|
||||
this.name = config.name;
|
||||
this.help = config.help;
|
||||
this.collect = config.collect;
|
||||
if (config.registers) {
|
||||
for (const r of config.registers) {
|
||||
r.registerMetric(this);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
set(labelsOrValue: Record<string, string> | number, value?: number): void {
|
||||
if (typeof labelsOrValue === 'number') {
|
||||
this.value = labelsOrValue;
|
||||
} else {
|
||||
const key = this.labelsToKey(labelsOrValue);
|
||||
this.labelledValues.set(key, value!);
|
||||
}
|
||||
}
|
||||
|
||||
inc(labelsOrAmount?: Record<string, string> | number, amount?: number): void {
|
||||
if (labelsOrAmount === undefined) {
|
||||
this.value += 1;
|
||||
} else if (typeof labelsOrAmount === 'number') {
|
||||
this.value += labelsOrAmount;
|
||||
} else {
|
||||
const key = this.labelsToKey(labelsOrAmount);
|
||||
const cur = this.labelledValues.get(key) || 0;
|
||||
this.labelledValues.set(key, cur + (amount ?? 1));
|
||||
}
|
||||
}
|
||||
|
||||
async getLines(): Promise<string[]> {
|
||||
const lines: string[] = [];
|
||||
if (this.labelledValues.size > 0) {
|
||||
for (const [key, val] of this.labelledValues) {
|
||||
lines.push(`${this.name}{${key}} ${formatValue(val)}`);
|
||||
}
|
||||
} else {
|
||||
lines.push(`${this.name} ${formatValue(this.value)}`);
|
||||
}
|
||||
return lines;
|
||||
}
|
||||
|
||||
/** Reset all values */
|
||||
reset(): void {
|
||||
this.value = 0;
|
||||
this.labelledValues.clear();
|
||||
}
|
||||
|
||||
private labelsToKey(labels: Record<string, string>): string {
|
||||
return Object.entries(labels)
|
||||
.map(([k, v]) => `${k}="${v}"`)
|
||||
.join(',');
|
||||
}
|
||||
}
|
||||
|
||||
// ── Counter ─────────────────────────────────────────────────────────────────
|
||||
|
||||
export class Counter implements IMetric {
|
||||
public name: string;
|
||||
public help: string;
|
||||
public type = 'counter';
|
||||
public collect?: () => void | Promise<void>;
|
||||
private value = 0;
|
||||
private labelledValues = new Map<string, number>();
|
||||
|
||||
constructor(config: ICounterConfig) {
|
||||
this.name = config.name;
|
||||
this.help = config.help;
|
||||
this.collect = config.collect;
|
||||
if (config.registers) {
|
||||
for (const r of config.registers) {
|
||||
r.registerMetric(this);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
inc(labelsOrAmount?: Record<string, string> | number, amount?: number): void {
|
||||
if (labelsOrAmount === undefined) {
|
||||
this.value += 1;
|
||||
} else if (typeof labelsOrAmount === 'number') {
|
||||
this.value += labelsOrAmount;
|
||||
} else {
|
||||
const key = this.labelsToKey(labelsOrAmount);
|
||||
const cur = this.labelledValues.get(key) || 0;
|
||||
this.labelledValues.set(key, cur + (amount ?? 1));
|
||||
}
|
||||
}
|
||||
|
||||
async getLines(): Promise<string[]> {
|
||||
const lines: string[] = [];
|
||||
if (this.labelledValues.size > 0) {
|
||||
for (const [key, val] of this.labelledValues) {
|
||||
lines.push(`${this.name}{${key}} ${formatValue(val)}`);
|
||||
}
|
||||
} else {
|
||||
lines.push(`${this.name} ${formatValue(this.value)}`);
|
||||
}
|
||||
return lines;
|
||||
}
|
||||
|
||||
reset(): void {
|
||||
this.value = 0;
|
||||
this.labelledValues.clear();
|
||||
}
|
||||
|
||||
private labelsToKey(labels: Record<string, string>): string {
|
||||
return Object.entries(labels)
|
||||
.map(([k, v]) => `${k}="${v}"`)
|
||||
.join(',');
|
||||
}
|
||||
}
|
||||
|
||||
// ── Histogram ───────────────────────────────────────────────────────────────
|
||||
|
||||
export class Histogram implements IMetric {
|
||||
public name: string;
|
||||
public help: string;
|
||||
public type = 'histogram';
|
||||
public collect?: () => void | Promise<void>;
|
||||
private bucketBounds: number[];
|
||||
private bucketCounts: number[];
|
||||
private sum = 0;
|
||||
private count = 0;
|
||||
private labelledData = new Map<
|
||||
string,
|
||||
{ bucketCounts: number[]; sum: number; count: number }
|
||||
>();
|
||||
|
||||
constructor(config: IHistogramConfig) {
|
||||
this.name = config.name;
|
||||
this.help = config.help;
|
||||
this.bucketBounds = config.buckets || [0.005, 0.01, 0.025, 0.05, 0.1, 0.25, 0.5, 1, 2.5, 5, 10];
|
||||
this.bucketCounts = new Array(this.bucketBounds.length).fill(0);
|
||||
this.collect = config.collect;
|
||||
if (config.registers) {
|
||||
for (const r of config.registers) {
|
||||
r.registerMetric(this);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
observe(labelsOrValue: Record<string, string> | number, value?: number): void {
|
||||
if (typeof labelsOrValue === 'number') {
|
||||
this.observeUnlabelled(labelsOrValue);
|
||||
} else {
|
||||
const key = this.labelsToKey(labelsOrValue);
|
||||
let data = this.labelledData.get(key);
|
||||
if (!data) {
|
||||
data = {
|
||||
bucketCounts: new Array(this.bucketBounds.length).fill(0),
|
||||
sum: 0,
|
||||
count: 0,
|
||||
};
|
||||
this.labelledData.set(key, data);
|
||||
}
|
||||
data.sum += value!;
|
||||
data.count += 1;
|
||||
for (let i = 0; i < this.bucketBounds.length; i++) {
|
||||
if (value! <= this.bucketBounds[i]) {
|
||||
data.bucketCounts[i]++;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
private observeUnlabelled(val: number): void {
|
||||
this.sum += val;
|
||||
this.count += 1;
|
||||
for (let i = 0; i < this.bucketBounds.length; i++) {
|
||||
if (val <= this.bucketBounds[i]) {
|
||||
this.bucketCounts[i]++;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
async getLines(): Promise<string[]> {
|
||||
const lines: string[] = [];
|
||||
if (this.labelledData.size > 0) {
|
||||
for (const [key, data] of this.labelledData) {
|
||||
for (let i = 0; i < this.bucketBounds.length; i++) {
|
||||
lines.push(
|
||||
`${this.name}_bucket{${key},le="${this.bucketBounds[i]}"} ${data.bucketCounts[i]}`
|
||||
);
|
||||
}
|
||||
lines.push(`${this.name}_bucket{${key},le="+Inf"} ${data.count}`);
|
||||
lines.push(`${this.name}_sum{${key}} ${formatValue(data.sum)}`);
|
||||
lines.push(`${this.name}_count{${key}} ${data.count}`);
|
||||
}
|
||||
} else {
|
||||
for (let i = 0; i < this.bucketBounds.length; i++) {
|
||||
lines.push(
|
||||
`${this.name}_bucket{le="${this.bucketBounds[i]}"} ${this.bucketCounts[i]}`
|
||||
);
|
||||
}
|
||||
lines.push(`${this.name}_bucket{le="+Inf"} ${this.count}`);
|
||||
lines.push(`${this.name}_sum ${formatValue(this.sum)}`);
|
||||
lines.push(`${this.name}_count ${this.count}`);
|
||||
}
|
||||
return lines;
|
||||
}
|
||||
|
||||
reset(): void {
|
||||
this.sum = 0;
|
||||
this.count = 0;
|
||||
this.bucketCounts.fill(0);
|
||||
this.labelledData.clear();
|
||||
}
|
||||
|
||||
private labelsToKey(labels: Record<string, string>): string {
|
||||
return Object.entries(labels)
|
||||
.map(([k, v]) => `${k}="${v}"`)
|
||||
.join(',');
|
||||
}
|
||||
}
|
||||
|
||||
// ── Default Metrics Collectors ──────────────────────────────────────────────
|
||||
|
||||
export function collectDefaultMetrics(registry: Registry): void {
|
||||
registerProcessCpuTotal(registry);
|
||||
registerProcessStartTime(registry);
|
||||
registerProcessMemory(registry);
|
||||
registerProcessOpenFds(registry);
|
||||
registerProcessMaxFds(registry);
|
||||
registerEventLoopLag(registry);
|
||||
registerProcessHandles(registry);
|
||||
registerProcessRequests(registry);
|
||||
registerProcessResources(registry);
|
||||
registerHeapSizeAndUsed(registry);
|
||||
registerHeapSpaces(registry);
|
||||
registerVersion(registry);
|
||||
registerGc(registry);
|
||||
}
|
||||
|
||||
function registerProcessCpuTotal(registry: Registry): void {
|
||||
const userGauge = new Gauge({
|
||||
name: 'process_cpu_user_seconds_total',
|
||||
help: 'Total user CPU time spent in seconds.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
const now = process.cpuUsage();
|
||||
userGauge.set(now.user / 1e6);
|
||||
},
|
||||
});
|
||||
|
||||
const systemGauge = new Gauge({
|
||||
name: 'process_cpu_system_seconds_total',
|
||||
help: 'Total system CPU time spent in seconds.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
const now = process.cpuUsage();
|
||||
systemGauge.set(now.system / 1e6);
|
||||
},
|
||||
});
|
||||
|
||||
new Gauge({
|
||||
name: 'process_cpu_seconds_total',
|
||||
help: 'Total user and system CPU time spent in seconds.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
const now = process.cpuUsage();
|
||||
this.set((now.user + now.system) / 1e6);
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerProcessStartTime(registry: Registry): void {
|
||||
const startTimeSeconds = Math.floor(Date.now() / 1000 - process.uptime());
|
||||
new Gauge({
|
||||
name: 'process_start_time_seconds',
|
||||
help: 'Start time of the process since unix epoch in seconds.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
this.set(startTimeSeconds);
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerProcessMemory(registry: Registry): void {
|
||||
new Gauge({
|
||||
name: 'process_resident_memory_bytes',
|
||||
help: 'Resident memory size in bytes.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
this.set(process.memoryUsage.rss());
|
||||
},
|
||||
});
|
||||
|
||||
new Gauge({
|
||||
name: 'process_virtual_memory_bytes',
|
||||
help: 'Virtual memory size in bytes.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
try {
|
||||
const status = fs.readFileSync('/proc/self/status', 'utf8');
|
||||
const match = status.match(/VmSize:\s+(\d+)\s+kB/);
|
||||
if (match) {
|
||||
this.set(parseInt(match[1], 10) * 1024);
|
||||
}
|
||||
} catch {
|
||||
// not on Linux — skip
|
||||
}
|
||||
},
|
||||
});
|
||||
|
||||
new Gauge({
|
||||
name: 'process_heap_bytes',
|
||||
help: 'Process heap size in bytes.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
this.set(process.memoryUsage().heapUsed);
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerProcessOpenFds(registry: Registry): void {
|
||||
new Gauge({
|
||||
name: 'process_open_fds',
|
||||
help: 'Number of open file descriptors.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
try {
|
||||
const fds = fs.readdirSync('/proc/self/fd');
|
||||
this.set(fds.length);
|
||||
} catch {
|
||||
this.set(0);
|
||||
}
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerProcessMaxFds(registry: Registry): void {
|
||||
new Gauge({
|
||||
name: 'process_max_fds',
|
||||
help: 'Maximum number of open file descriptors.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
try {
|
||||
const limits = fs.readFileSync('/proc/self/limits', 'utf8');
|
||||
const match = limits.match(/Max open files\s+(\d+)/);
|
||||
if (match) {
|
||||
this.set(parseInt(match[1], 10));
|
||||
}
|
||||
} catch {
|
||||
this.set(0);
|
||||
}
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerEventLoopLag(registry: Registry): void {
|
||||
let histogram: ReturnType<typeof monitorEventLoopDelay> | null = null;
|
||||
try {
|
||||
histogram = monitorEventLoopDelay({ resolution: 10 });
|
||||
histogram.enable();
|
||||
} catch {
|
||||
// Not available in this runtime
|
||||
}
|
||||
|
||||
new Gauge({
|
||||
name: 'nodejs_eventloop_lag_seconds',
|
||||
help: 'Lag of event loop in seconds.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
if (histogram) {
|
||||
this.set(histogram.mean / 1e9);
|
||||
}
|
||||
},
|
||||
});
|
||||
|
||||
new Gauge({
|
||||
name: 'nodejs_eventloop_lag_min_seconds',
|
||||
help: 'The minimum recorded event loop delay.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
if (histogram) {
|
||||
this.set(histogram.min / 1e9);
|
||||
}
|
||||
},
|
||||
});
|
||||
|
||||
new Gauge({
|
||||
name: 'nodejs_eventloop_lag_max_seconds',
|
||||
help: 'The maximum recorded event loop delay.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
if (histogram) {
|
||||
this.set(histogram.max / 1e9);
|
||||
}
|
||||
},
|
||||
});
|
||||
|
||||
new Gauge({
|
||||
name: 'nodejs_eventloop_lag_mean_seconds',
|
||||
help: 'The mean of the recorded event loop delays.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
if (histogram) {
|
||||
this.set(histogram.mean / 1e9);
|
||||
}
|
||||
},
|
||||
});
|
||||
|
||||
new Gauge({
|
||||
name: 'nodejs_eventloop_lag_stddev_seconds',
|
||||
help: 'The standard deviation of the recorded event loop delays.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
if (histogram) {
|
||||
this.set(histogram.stddev / 1e9);
|
||||
}
|
||||
},
|
||||
});
|
||||
|
||||
for (const p of [50, 90, 99]) {
|
||||
new Gauge({
|
||||
name: `nodejs_eventloop_lag_p${p}_seconds`,
|
||||
help: `The ${p}th percentile of the recorded event loop delays.`,
|
||||
registers: [registry],
|
||||
collect() {
|
||||
if (histogram) {
|
||||
this.set(histogram.percentile(p) / 1e9);
|
||||
}
|
||||
},
|
||||
});
|
||||
}
|
||||
}
|
||||
|
||||
function registerProcessHandles(registry: Registry): void {
|
||||
new Gauge({
|
||||
name: 'nodejs_active_handles_total',
|
||||
help: 'Number of active libuv handles grouped by handle type.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
const handles = (process as any)._getActiveHandles?.();
|
||||
this.set(handles ? handles.length : 0);
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerProcessRequests(registry: Registry): void {
|
||||
new Gauge({
|
||||
name: 'nodejs_active_requests_total',
|
||||
help: 'Number of active libuv requests grouped by request type.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
const requests = (process as any)._getActiveRequests?.();
|
||||
this.set(requests ? requests.length : 0);
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerProcessResources(registry: Registry): void {
|
||||
new Gauge({
|
||||
name: 'nodejs_active_resources_total',
|
||||
help: 'Number of active resources that are currently keeping the event loop alive.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
try {
|
||||
const resources = (process as any).getActiveResourcesInfo?.();
|
||||
this.set(resources ? resources.length : 0);
|
||||
} catch {
|
||||
this.set(0);
|
||||
}
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerHeapSizeAndUsed(registry: Registry): void {
|
||||
new Gauge({
|
||||
name: 'nodejs_heap_size_total_bytes',
|
||||
help: 'Process heap size from Node.js in bytes.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
this.set(process.memoryUsage().heapTotal);
|
||||
},
|
||||
});
|
||||
|
||||
new Gauge({
|
||||
name: 'nodejs_heap_size_used_bytes',
|
||||
help: 'Process heap size used from Node.js in bytes.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
this.set(process.memoryUsage().heapUsed);
|
||||
},
|
||||
});
|
||||
|
||||
new Gauge({
|
||||
name: 'nodejs_external_memory_bytes',
|
||||
help: 'Node.js external memory size in bytes.',
|
||||
registers: [registry],
|
||||
collect() {
|
||||
this.set(process.memoryUsage().external);
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerHeapSpaces(registry: Registry): void {
|
||||
const spaceGauge = new Gauge({
|
||||
name: 'nodejs_heap_space_size_total_bytes',
|
||||
help: 'Process heap space size total from Node.js in bytes.',
|
||||
labelNames: ['space'],
|
||||
registers: [registry],
|
||||
collect() {
|
||||
spaceGauge.reset();
|
||||
const spaces = v8.getHeapSpaceStatistics();
|
||||
for (const space of spaces) {
|
||||
spaceGauge.set({ space: space.space_name }, space.space_size);
|
||||
}
|
||||
},
|
||||
});
|
||||
|
||||
const usedGauge = new Gauge({
|
||||
name: 'nodejs_heap_space_size_used_bytes',
|
||||
help: 'Process heap space size used from Node.js in bytes.',
|
||||
labelNames: ['space'],
|
||||
registers: [registry],
|
||||
collect() {
|
||||
usedGauge.reset();
|
||||
const spaces = v8.getHeapSpaceStatistics();
|
||||
for (const space of spaces) {
|
||||
usedGauge.set({ space: space.space_name }, space.space_used_size);
|
||||
}
|
||||
},
|
||||
});
|
||||
|
||||
const availableGauge = new Gauge({
|
||||
name: 'nodejs_heap_space_size_available_bytes',
|
||||
help: 'Process heap space size available from Node.js in bytes.',
|
||||
labelNames: ['space'],
|
||||
registers: [registry],
|
||||
collect() {
|
||||
availableGauge.reset();
|
||||
const spaces = v8.getHeapSpaceStatistics();
|
||||
for (const space of spaces) {
|
||||
availableGauge.set({ space: space.space_name }, space.space_available_size);
|
||||
}
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerVersion(registry: Registry): void {
|
||||
const versionParts = process.version.slice(1).split('.').map(Number);
|
||||
const gauge = new Gauge({
|
||||
name: 'nodejs_version_info',
|
||||
help: 'Node.js version info.',
|
||||
labelNames: ['version', 'major', 'minor', 'patch'],
|
||||
registers: [registry],
|
||||
collect() {
|
||||
gauge.set(
|
||||
{
|
||||
version: process.version,
|
||||
major: String(versionParts[0]),
|
||||
minor: String(versionParts[1]),
|
||||
patch: String(versionParts[2]),
|
||||
},
|
||||
1
|
||||
);
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
function registerGc(registry: Registry): void {
|
||||
const gcHistogram = new Histogram({
|
||||
name: 'nodejs_gc_duration_seconds',
|
||||
help: 'Garbage collection duration by kind, in seconds.',
|
||||
labelNames: ['kind'],
|
||||
buckets: [0.001, 0.01, 0.1, 1, 2, 5],
|
||||
registers: [registry],
|
||||
});
|
||||
|
||||
const kindLabels: Record<number, string> = {
|
||||
1: 'Scavenge',
|
||||
2: 'Mark/Sweep/Compact',
|
||||
4: 'IncrementalMarking',
|
||||
8: 'ProcessWeakCallbacks',
|
||||
15: 'All',
|
||||
};
|
||||
|
||||
try {
|
||||
const obs = new PerformanceObserver((list) => {
|
||||
for (const entry of list.getEntries()) {
|
||||
const gcEntry = entry as any;
|
||||
const kind = kindLabels[gcEntry.detail?.kind ?? gcEntry.kind] || 'Unknown';
|
||||
gcHistogram.observe({ kind }, entry.duration / 1000);
|
||||
}
|
||||
});
|
||||
obs.observe({ entryTypes: ['gc'] });
|
||||
} catch {
|
||||
// GC observation not available
|
||||
}
|
||||
}
|
||||
|
||||
// ── Helpers ─────────────────────────────────────────────────────────────────
|
||||
|
||||
function formatValue(v: number): string {
|
||||
if (Number.isInteger(v)) return String(v);
|
||||
return v.toString();
|
||||
}
|
||||
94
ts/smartmetrics.sysusage.ts
Normal file
94
ts/smartmetrics.sysusage.ts
Normal file
@@ -0,0 +1,94 @@
|
||||
import * as fs from 'fs';
|
||||
import * as os from 'os';
|
||||
|
||||
export interface ISystemUsageResult {
|
||||
cpuPercent: number; // 0-100% system-wide CPU utilization
|
||||
memTotalBytes: number; // total physical RAM
|
||||
memAvailableBytes: number; // available memory
|
||||
memUsedBytes: number; // memTotal - memAvailable
|
||||
memUsedPercent: number; // 0-100%
|
||||
loadAvg1: number; // 1-min load average
|
||||
loadAvg5: number; // 5-min load average
|
||||
loadAvg15: number; // 15-min load average
|
||||
}
|
||||
|
||||
// History for system CPU delta tracking
|
||||
interface ICpuSnapshot {
|
||||
idle: number;
|
||||
total: number;
|
||||
}
|
||||
|
||||
let prevCpuSnapshot: ICpuSnapshot | null = null;
|
||||
|
||||
function readProcStat(): ICpuSnapshot | null {
|
||||
try {
|
||||
const content = fs.readFileSync('/proc/stat', 'utf8');
|
||||
const firstLine = content.split('\n')[0]; // "cpu user nice system idle iowait irq softirq steal ..."
|
||||
const parts = firstLine.split(/\s+/).slice(1).map(Number);
|
||||
// parts: [user, nice, system, idle, iowait, irq, softirq, steal, ...]
|
||||
const idle = parts[3] + (parts[4] || 0); // idle + iowait
|
||||
const total = parts.reduce((sum, v) => sum + v, 0);
|
||||
return { idle, total };
|
||||
} catch {
|
||||
return null;
|
||||
}
|
||||
}
|
||||
|
||||
function getMemoryInfo(): { totalBytes: number; availableBytes: number } {
|
||||
try {
|
||||
const content = fs.readFileSync('/proc/meminfo', 'utf8');
|
||||
let memTotal = 0;
|
||||
let memAvailable = 0;
|
||||
for (const line of content.split('\n')) {
|
||||
if (line.startsWith('MemTotal:')) {
|
||||
memTotal = parseInt(line.split(/\s+/)[1], 10) * 1024; // kB to bytes
|
||||
} else if (line.startsWith('MemAvailable:')) {
|
||||
memAvailable = parseInt(line.split(/\s+/)[1], 10) * 1024;
|
||||
}
|
||||
}
|
||||
if (memTotal > 0 && memAvailable > 0) {
|
||||
return { totalBytes: memTotal, availableBytes: memAvailable };
|
||||
}
|
||||
} catch {
|
||||
// fall through to os fallback
|
||||
}
|
||||
// Fallback using os module
|
||||
const totalBytes = os.totalmem();
|
||||
const availableBytes = os.freemem();
|
||||
return { totalBytes, availableBytes };
|
||||
}
|
||||
|
||||
export async function getSystemUsage(): Promise<ISystemUsageResult> {
|
||||
// CPU
|
||||
let cpuPercent = 0;
|
||||
const currentSnapshot = readProcStat();
|
||||
if (currentSnapshot && prevCpuSnapshot) {
|
||||
const totalDelta = currentSnapshot.total - prevCpuSnapshot.total;
|
||||
const idleDelta = currentSnapshot.idle - prevCpuSnapshot.idle;
|
||||
if (totalDelta > 0) {
|
||||
cpuPercent = ((totalDelta - idleDelta) / totalDelta) * 100;
|
||||
}
|
||||
}
|
||||
if (currentSnapshot) {
|
||||
prevCpuSnapshot = currentSnapshot;
|
||||
}
|
||||
|
||||
// Memory
|
||||
const mem = getMemoryInfo();
|
||||
const memUsedBytes = mem.totalBytes - mem.availableBytes;
|
||||
const memUsedPercent = mem.totalBytes > 0 ? (memUsedBytes / mem.totalBytes) * 100 : 0;
|
||||
|
||||
// Load averages
|
||||
const [loadAvg1, loadAvg5, loadAvg15] = os.loadavg();
|
||||
|
||||
return {
|
||||
cpuPercent,
|
||||
memTotalBytes: mem.totalBytes,
|
||||
memAvailableBytes: mem.availableBytes,
|
||||
memUsedBytes,
|
||||
memUsedPercent,
|
||||
loadAvg1,
|
||||
loadAvg5,
|
||||
loadAvg15,
|
||||
};
|
||||
}
|
||||
Reference in New Issue
Block a user