Skip to content

Crawlee—A web scraping and browser automation library for Node.js to build reliable crawlers. In JavaScript and TypeScript. Extract data for AI, LLMs, RAG, or GPTs. Download HTML, PDF, JPG, PNG, and other files from websites. Works with Puppeteer, Playwright, Cheerio, JSDOM, and raw HTTP. Both headful and headless mode. With proxy rotation.

Kill browser process on pipe errors

Ondra Urbancc14a1224817
diff --git a/src/puppeteer_pool.js b/src/puppeteer_pool.js
index 674927d5f..c2904527e 100644
--- a/src/puppeteer_pool.js
+++ b/src/puppeteer_pool.js
@@ -336,16 +336,16 @@ class PuppeteerPool {
* When using "pipe: true" launch option, the pipe can break for various reasons,
* typically when the browser gets killed by PuppeteerPool. We swallow those errors,
* because there's nothing we can do anyway and we don't want them to crash Node.
- * @param browserProcess
+ * @param chromeProcess
* @ignore
*/
- _attachPipeErrorHandlers(browserProcess) { // eslint-disable-line class-methods-use-this
- const writeSocket = browserProcess.stdio[PUPPETEER_PIPE_TRANSPORT_WRITE];
- const readSocket = browserProcess.stdio[PUPPETEER_PIPE_TRANSPORT_READ];
+ _attachPipeErrorHandlers(chromeProcess) { // eslint-disable-line class-methods-use-this
+ const writeSocket = chromeProcess.stdio[PUPPETEER_PIPE_TRANSPORT_WRITE];
+ const readSocket = chromeProcess.stdio[PUPPETEER_PIPE_TRANSPORT_READ];
const errorHandler = (err) => {
- if (err.code === 'EPIPE') log.debug('Puppeteer Pool: Browser connection failed with an EPIPE error.');
- else throw err;
+ log.debug('Puppeteer Pool: Browser connection failed. Browser will shut down.', { message: err.message });
+ if (chromeProcess) chromeProcess.kill('SIGKILL');
};
writeSocket.on('error', errorHandler);