you need paid worker plan

C

cloudkite•7/16/24, 8:48 AM

I have set keep_alive option to 10mins, however my browser session is closed after 10mins even if I have had active connections within the 10min timeframe. Shouldn't idle termination only happen if browser has been idle for 10mins since the most recent connection? At the moment it appears that browser is closed 10mins from the first connection

Aakazwz you need paid worker plan

A

Arvind•7/16/24, 4:23 PM

like i dont know if i have to pay for it but i need access, whether its paid or free

Ssc Where did you sign up for it? The beta access has ended. Right now, you only nee...

A

Arvind•7/16/24, 6:27 PM

Thank you for sharing the link. The Beta access URL was found somewhere on Reddit I guess.

Anyways, When I click on the URL you provided, and signup for a paid plan (which is $5 per month), I get an error that "I don't have access to that page" and then I get redirected to the worker's page. There I don't see any way to upgrade or get on the paid plan

A

Arvind•7/16/24, 6:27 PM

Screenshot_2024-07-16_at_11.54.32_PM.png

Screenshot_2024-07-16_at_11.54.47_PM.png

Sscythe I think that after the 10 mins pass, you’re back at the idle time of 1 min. So f...

C

cloudkite•7/16/24, 7:29 PM

Any idea if thats intentional or a bug? If intentional seems like a very confusing API choice

C

cloudkite•7/16/24, 8:29 PM

Have submitted a bug with a minimal repo see https://github.com/cloudflare/puppeteer/issues/53

GitHub

[Bug]: Browser is closed despite having set keep_alive setting to 1...

Bug description Steps to reproduce the problem: Minimal reproduction see https://github.com/cloudkite/keep-alive-test. Have got a simple worker which crawls https://example.com every 5mins. Browser...

P

Puliczek•7/17/24, 3:57 PM

How to make Browser rendering working in local env? Can it handle like 100+ users requests per minute? I need only to scrap pages.

Sscythe Are you using @types/node >=20.10 ? Had that issue with 20.8 and an update fixe...

D

dave•7/21/24, 12:03 AM

I did

yarn add @types/node -D

yarn add @types/node -D

yarn add @types/node -D

yarn add @types/node -D, but I'm still getting this error

Ddave I did `yarn add @types/node -D`, but I'm still getting this error

J

James•7/21/24, 3:48 AM

You might be hitting https://github.com/cloudflare/workerd/issues/1298

D

dave•7/21/24, 3:50 AM

all good, I ended up switching my project away from browser rendering.

몰

몰라세스•7/26/24, 4:31 AM

It seems like you can use a string that is unlikely to be included in the URL, and then divide it by that string on the server.

For example, if divide is , (comma), you can divide the string A "a.com,b.com,c.com" with A.split(",").

몰

몰라세스•7/26/24, 4:32 AM

I'm not sure since I haven't looked at the source code in detail, but this is the simplest way to pass an Array via URL.

몰몰라세스 It seems like you can use a string that is unlikely to be included in the URL, a...

H

Hamed•7/26/24, 4:58 AM

I'm not sure really how the

split

split

split

split method could be helpful in this case. But appreciate your comment.

H

Hamed•7/28/24, 1:15 PM

Hi again!

I've got this Worker already deployed to the Cloudflare. I need to run it every

minutes. So I did it under

Settings > Triggers > Cron Triggers

Settings > Triggers > Cron Triggers

Settings > Triggers > Cron Triggers

Settings > Triggers > Cron Triggers of that specific worker (picture). But the worker never triggered and never run? And I can see the errors (picture) but I don't know where to track down the logs! I appreciate any thought!?

import puppeteer from '@cloudflare/puppeteer';
import { nanoid } from 'nanoid';

const URLS = [
    'https://cloudflare.com',
    'https://google.com',
    'https://wikipedia.org',
];

export default {
    async fetch(_, env) {
        let browser;
        try {
            browser = await puppeteer.launch(env.MYBROWSER);

            const processUrl = async (url) => {
                const page = await browser.newPage();
                await page.goto(url);
                let img = await page.screenshot();

                const filename = new URL(url).hostname.replace(/^www\./, '').replace(/\./g, '-');
                const id = nanoid(5);

                await env.BROWSER_KV_DEMO.put(`${filename}_${id}`, img);

                await page.close();
            };

            await Promise.all(URLS.map(processUrl));
        } catch (error) {
            console.error({ error });
        } finally {
            if (browser) {
                await browser.close();
            }
        }
        return new Response('ok');
    },
};

import puppeteer from '@cloudflare/puppeteer';
import { nanoid } from 'nanoid';

const URLS = [
    'https://cloudflare.com',
    'https://google.com',
    'https://wikipedia.org',
];

export default {
    async fetch(_, env) {
        let browser;
        try {
            browser = await puppeteer.launch(env.MYBROWSER);

            const processUrl = async (url) => {
                const page = await browser.newPage();
                await page.goto(url);
                let img = await page.screenshot();

                const filename = new URL(url).hostname.replace(/^www\./, '').replace(/\./g, '-');
                const id = nanoid(5);

                await env.BROWSER_KV_DEMO.put(`${filename}_${id}`, img);

                await page.close();
            };

            await Promise.all(URLS.map(processUrl));
        } catch (error) {
            console.error({ error });
        } finally {
            if (browser) {
                await browser.close();
            }
        }
        return new Response('ok');
    },
};

import puppeteer from '@cloudflare/puppeteer';
import { nanoid } from 'nanoid';

const URLS = [
    'https://cloudflare.com',
    'https://google.com',
    'https://wikipedia.org',
];

export default {
    async fetch(_, env) {
        let browser;
        try {
            browser = await puppeteer.launch(env.MYBROWSER);

            const processUrl = async (url) => {
                const page = await browser.newPage();
                await page.goto(url);
                let img = await page.screenshot();

                const filename = new URL(url).hostname.replace(/^www\./, '').replace(/\./g, '-');
                const id = nanoid(5);

                await env.BROWSER_KV_DEMO.put(`${filename}_${id}`, img);

                await page.close();
            };

            await Promise.all(URLS.map(processUrl));
        } catch (error) {
            console.error({ error });
        } finally {
            if (browser) {
                await browser.close();
            }
        }
        return new Response('ok');
    },
};

import puppeteer from '@cloudflare/puppeteer';
import { nanoid } from 'nanoid';

const URLS = [
    'https://cloudflare.com',
    'https://google.com',
    'https://wikipedia.org',
];

export default {
    async fetch(_, env) {
        let browser;
        try {
            browser = await puppeteer.launch(env.MYBROWSER);

            const processUrl = async (url) => {
                const page = await browser.newPage();
                await page.goto(url);
                let img = await page.screenshot();

                const filename = new URL(url).hostname.replace(/^www\./, '').replace(/\./g, '-');
                const id = nanoid(5);

                await env.BROWSER_KV_DEMO.put(`${filename}_${id}`, img);

                await page.close();
            };

            await Promise.all(URLS.map(processUrl));
        } catch (error) {
            console.error({ error });
        } finally {
            if (browser) {
                await browser.close();
            }
        }
        return new Response('ok');
    },
};

HHamed Hi again! 👋 I've got this Worker already deployed to the Cloudflare. I need ...

K

kian•7/28/24, 1:38 PM

You need a

scheduled

scheduled

scheduled

scheduled handler,

fetch

fetch

is for HTTP requests

K

kian•7/28/24, 1:38 PM

https://developers.cloudflare.com/workers/configuration/cron-triggers/#1-define-a-scheduled-event-listener

Y

yangyangy•7/28/24, 3:52 PM

workers.api.error.no_access_to_browser_worker [code: 10088]

J

jjjrmy•7/28/24, 9:28 PM

is there any way to get around the Cloudflare Human Check?

Jjjjrmy is there any way to get around the Cloudflare Human Check?

H

Hello, I’m Allie!•7/28/24, 9:52 PM

Be a human, click the button?

H

Hello, I’m Allie!•7/28/24, 9:54 PM

Oh, in Browser Rendering? Not that I know of

H

Hello, I’m Allie!•7/28/24, 9:54 PM

Unless it is your website

E

Erisa•7/28/24, 10:26 PM

Well browser rendering is not a human, so it would be intentional that you can't pass it.

E

Erisa•7/28/24, 10:26 PM

If it's your website you can identify those requests in some way and bypass challenges for them

E

emil•8/2/24, 12:06 PM

Not exactly sure on the concurrent sessions pricing. If I run 10 sessions at a time max is it correct I won't ever be billed for concurrent sessions? (I'm on Workers paid)

E

emil•8/2/24, 12:07 PM

Also I have a 1/minute Cron trigger on the same worker with my Browser binding, but it doesn't actually use the browser binding, so that shouldn't start any sessions, correct?

E

emil•8/2/24, 12:51 PM

Also getting this on page.pdf() calls with Puppeteer - looks like the GitHub issue on it had no answers since April: https://github.com/cloudflare/puppeteer/issues/37

GitHub

[Bug]: Puppeteer PDF generation via Page.pdf() method · Issue #37 ·...

Bug description I was creating a simple html to pdf converter when I stumbled into a little issue with all the current version of the fork. In the current implementation of the Page.pdf method in t...

0

0xbrmn•8/11/24, 1:23 PM

Adding WebGL to Browser rendering would be amazing.

M

Mr. Technician•8/12/24, 2:03 PM

Are there plans to allow more than 2 browser per minute for the rendering API?

M

Mr. Technician•8/12/24, 2:03 PM

it's an unfortunate limitation that means I can't use it for my project

MMr. Technician Are there plans to allow more than 2 browser per minute for the rendering API?

P

Peps•8/12/24, 2:30 PM

Hm out of curiosity, is there a reason you can't re-use sessions?

P

Peps•8/12/24, 2:31 PM

There's a guide on how to re-use session at https://developers.cloudflare.com/browser-rendering/get-started/reuse-sessions/

Cloudflare Docs

Reuse sessions · Browser Rendering docs

The best way to improve the performance of your browser rendering worker is to reuse sessions. One way to do that is via Durable Objects, which allows …

M

Mr. Technician•8/12/24, 2:31 PM

hmm I was not aware of that as an option...

M

Mr. Technician•8/12/24, 2:31 PM

but I would still need to run tasks in serial, no?

M

Mr. Technician•8/12/24, 2:33 PM

I have site that currently uses Cloudflare pages for the React piece but uses Azure Functions coupled with Browserless in a VM for web scraping

M

Mr. Technician•8/12/24, 2:52 PM

the issue is I need to spin up a new browser each time a user makes a new request. Reuse is on the table but not if I need many concurrent browsers

P

Peps•8/12/24, 3:21 PM

I need to spin up a new browser each time

why?

PPeps > I need to spin up a new browser each time why?

M

Mr. Technician•8/12/24, 7:32 PM

I guess I am using the wrong term. What I need is multiple discrete puppeteer sessions if there happen to be multiple concurrent users running different requests

M

Mr. Technician•8/12/24, 8:08 PM

https://developers.cloudflare.com/browser-rendering/platform/limits/

Cloudflare Docs

Limits · Browser Rendering docs

Control and interact with a headless browser instance programmatically.

M

Mr. Technician•8/12/24, 8:08 PM

"two concurrent browsers"

M

Mr. Technician•8/12/24, 8:09 PM

so does browser = session?

Similar Threads

Similar Threads

Similar Threads