Puppeteer-Pool is a lightweight and efficient library for managing multiple Puppeteer contexts with ease. Designed for scalable web scraping and automation, it helps you handle concurrency and resource management without the hassle.
-
npmnpm i puppeteer @hoplin/puppeteer-pool -
yarnyarn add puppeteer @hoplin/puppeteer-pool -
pnpmpnpm install puppeteer @hoplin/puppeteer-pool
-
Garbage collection is executed at the end of each context task
-
Client's method are now all 'static'. You don't need to create instance of client and need migration if you are using previous version(
< 2.1.0).getPoolMetricsrunTaskstopcheckInstanceInitalized
-
Add Priority Queue for task queue
-
taskQueueTypeoption is added toPuppeteerPoolStartOptions-
Priorty can be specified in the second factor of 'PuppeteerPool.runTask'. The larger the number, the higher the priority, and the priority is ignored in the Default Queue.
-
Enum:
QueueMode -
Example
import { ContextMode, PuppeteerPool, QueueMode, } from '@hoplin/puppeteer-pool'; const priority = 10; await PuppeteerPool.start({ taskQueueType: QueueMode.PRIORITY, }); PuppeteerPool.runTask(async (page) => { await page.goto(url); const title = await page.title(); return title; }, priority);
-
-
[ Client API ]
-
PuppeteePool
PuppeteerPoolis singleton class. You can usePuppeteerPool.startto initialize pool manager.
-
PuppeteerPool.start
-
Static Method
-
Description: Initialize pool manager. You need to call this function to start puppeteer pool. Even if you invoke this function multiple times with differenct arguments, it will return the first initialized instance.
-
Args:
PuppeteerPoolStartOptionstype PuppeteerPoolStartOptions = { /** * Number of concurrency, * Default is 3 */ concurrencyLevel: number; /** * Context mode * Default is ContextMode.SHARED */ contextMode: ContextMode; /** * Puppeteer launch options * Default is {} */ options?: puppeteer.LaunchOptions; /** * Custom config path */ customConfigPath?: string; /** * Enable log * Default is true */ enableLog?: boolean; /** * Log level * Default is LogLevel.DEBUG */ logLevel?: LogLevel; };
-
Return
Promise<PuppeteerPool>- Returns PuppeteerPool Instance.
-
-
PuppeteerPool.stop
- Static Method
- Description: Stop pool manager. It will close all sessions and terminate pool manager.
- Return
Promise<void>
-
PuppeteerPool.runTask
- Static Method
- Description: Run task in pool manager. It will return result of task.
- Args
- task
- Required
- Function
- Return
Promise<any>- Returns result of task(Same return type with task callback return type)
-
PuppeteerPool.getPoolMetrics
- Static Method
- Description: Get pool metrics. It will return metrics of pool manager.
- Return
{ memoryUsageValue: (Memory Usage in MegaBytes), memoryUsagePercentage: (Memory Usage with percentage), cpuUsage: (CPU Usage with percentage) }
import { ContextMode, PuppeteerPool, QueueMode } from '@hoplin/puppeteer-pool';
async function main() {
await PuppeteerPool.start({
concurrencyLevel: 6,
contextMode: ContextMode.ISOLATED,
customConfigPath: `./puppeteer-pool-config.json`,
taskQueueType: QueueMode.PRIORITY,
});
const baseUrls = [
'https://www.google.com',
'https://www.bing.com',
'https://github.com',
'https://www.naver.com',
'https://www.daum.net',
'https://www.youtube.com',
'https://www.amazon.com',
'https://www.netflix.com',
];
const urls = Array.from({ length: 50 }, (_, index) => {
const baseUrl = baseUrls[index % baseUrls.length];
const priority = Math.floor(Math.random() * 10) + 1;
return { url: baseUrl, priority: priority };
});
const promises = urls.map(({ url, priority }) => {
console.log(`Enqueue task: ${url}`);
return PuppeteerPool.runTask(async (page) => {
console.log(`Process task for ${url} with priority ${priority}`);
await page.goto(url);
const title = await page.title();
return title;
}, priority);
});
const titles = await Promise.all(promises);
console.log('[ Result length ] :', titles.length);
console.log('[ Expected Tasks ] :', urls.length);
console.log('[ Metrics ] :', await PuppeteerPool.getPoolMetrics());
}
main();- Pool Managing
- Config
- Support config customize
- Threshold Watcher
- Memory
- Support safe pool instance reset in runtime
- Metrics
- Support Metric by pool
- CPU usage of pool
- Memory usage of pool
- Managing session count in runtime
- Support Metric by pool
Default config should be puppeteer-pool-config.json in root directory path.
If config file are not given or invalid path, manager will use default defined configurations. Or if you want to pass
config path, you can pass path to start() function as parameter.
{
"session_pool": {
"width": 1080,
"height": 1024
},
"context": {
"timeout": 10
},
"threshold": {
"activate": true,
"interval": 5,
"memory": 2048
}
}width: Width of session poolheight: Height of session pool- Inteager Validation
widthshould be larger or equal than 50heightshould be larger or equal than 50widthshould be integerheightshould be integer
- Inteager Validation
timeout: Timeout of context.- Inteager Validation
timeoutshould be at least 1timeoutshould be integer
- Inteager Validation
activate: Activate threshold watcherinterval: Interval of threshold watchermemory: Memory threshold value- Inteager Validation
intervalshould be at least 1intervalshould be integermemoryshould be at least 1memoryshould be integer
- Inteager Validation
