web-ui
web-ui copied to clipboard
Own chrome browser not running
I try using my own browser, I set this up in my .env file:
CHROME_PATH="C:\Program Files\Google\Chrome\Application\chrome.exe" CHROME_USER_DATA="C:\Users\YourUsername\AppData\Local\Google\Chrome\User Data"
It open my chrome, but after then not running task at all, and no error occurred, just stuck in there.
I assume it may be due to multiple profile I have so I deleted all others but with main profile, but it still not working.
me too!
Set CHROME_USER_DATA empty, try again
Set CHROME_USER_DATA empty, try again
Now CHROME_USER_DATA is empty, an error occurs after running:
[src.browser.custom_browser] Failed to start a new Chrome instance.: BrowserType.connect_over_cdp: connect ECONNREFUSED ::1:9222
Call log:
- <ws preparing> retrieving websocket url from http://localhost:9222
ERROR [browser] Failed to initialize Playwright browser: To start chrome in Debug mode, you need to close all existing Chrome instances and try again otherwise we can not connect to the instance.
WARNING [src.agent.custom_agent] No history to create GIF from
and Chrome is closed.
Not working as well, the chrome do nothing and about a minute later it open the default one without my data and credentials.
Not working as well, it open my browser and not processing, then after a minute it open the default one and do the task, the one without my data
Check if there is a process occupying port 9222. If you are using Windows, enter the following command in the terminal: netstat -aon|findstr 9222. Then, run taskkill /f /pid {pid of 9222}.
For me using own browser was working yesterday and today morning perfectly. But suddenly this error started to appear:
INFO [src.agent.custom_agent]
📍 Step 1
ERROR [browser] Failed to update state: Page.screenshot: Protocol error (Page.captureScreenshot): Cannot take screenshot with 0 width.
Call log:
- taking page screenshot
- - disabled all CSS animations
- - waiting for fonts to load...
- - fonts loaded
ERROR [agent] ❌ Result failed 3/3 times:
Page.screenshot: Protocol error (Page.captureScreenshot): Cannot take screenshot with 0 width.
Call log:
- taking page screenshot
- - disabled all CSS animations
- - waiting for fonts to load...
- - fonts loaded
ERROR [agent] ❌ Stopping due to 3 consecutive failures
WARNING [src.agent.custom_agent] No history to create GIF from
I tried to install project again but it's still occurring. Chrome opens up but nothing else happens.
I have a same error, last week works fine, but today not working:
INFO [browser_use] BrowserUse logging setup complete with level info INFO [root] Anonymized telemetry enabled. See https://github.com/browser-use/browser-use for more information. INFO [agent] 🚀 Starting task: Search for Tour Agencies INFO [agent] 📍 Step 1 ERROR [browser] Failed to start a new Chrome instance.: BrowserType.connect_over_cdp: connect ECONNREFUSED ::1:9222 Call log:
retrieving websocket url from http://localhost:9222
WARNING [browser] Page load failed, continuing... ERROR [browser] Failed to start a new Chrome instance.: BrowserType.connect_over_cdp: connect ECONNREFUSED ::1:9222 Call log:
retrieving websocket url from http://localhost:9222
ERROR [agent] ❌ Result failed 1/3 times: To start chrome in Debug mode, you need to close all existing Chrome instances and try again otherwise we can not connect to the instance. INFO [agent]
Did anyone find the solution for this issue?
In case it helps anyone (in windows), I created a .bat file, like so: @echo off "C:\Program Files (x86)\Google\Chrome\Application\chrome.exe" --profile-directory="Profile 16" %*
And then set my path to CHROME_PATH=C:\Users\user\web-ui\my-chrome.bat
This way I'm able to have it load the exact profile I want to load.
In case it helps anyone (in windows), I created a .bat file, like so: @echo off "C:\Program Files (x86)\Google\Chrome\Application\chrome.exe" --profile-directory="Profile 16" %*
And then set my path to CHROME_PATH=C:\Users\user\web-ui\my-chrome.bat
This way I'm able to have it load the exact profile I want to load.
is the profile 16fixed or actually a variable
For me using own browser was working yesterday and today morning perfectly. But suddenly this error started to appear:
INFO [src.agent.custom_agent] 📍 Step 1 ERROR [browser] Failed to update state: Page.screenshot: Protocol error (Page.captureScreenshot): Cannot take screenshot with 0 width. Call log: - taking page screenshot - - disabled all CSS animations - - waiting for fonts to load... - - fonts loaded ERROR [agent] ❌ Result failed 3/3 times: Page.screenshot: Protocol error (Page.captureScreenshot): Cannot take screenshot with 0 width. Call log: - taking page screenshot - - disabled all CSS animations - - waiting for fonts to load... - - fonts loaded ERROR [agent] ❌ Stopping due to 3 consecutive failures WARNING [src.agent.custom_agent] No history to create GIF fromI tried to install project again but it's still occurring. Chrome opens up but nothing else happens.
@imlastrebor have you sorted this? I seem to have the same problem
what do i do as a linux user
I have a same error, last week works fine, but today not working:
INFO [browser_use] BrowserUse logging setup complete with level info INFO [root] Anonymized telemetry enabled. See https://github.com/browser-use/browser-use for more information. INFO [agent] 🚀 Starting task: Search for Tour Agencies INFO [agent] 📍 Step 1 ERROR [browser] Failed to start a new Chrome instance.: BrowserType.connect_over_cdp: connect ECONNREFUSED ::1:9222 Call log:
- retrieving websocket url from http://localhost:9222
WARNING [browser] Page load failed, continuing... ERROR [browser] Failed to start a new Chrome instance.: BrowserType.connect_over_cdp: connect ECONNREFUSED ::1:9222 Call log:
- retrieving websocket url from http://localhost:9222
ERROR [agent] ❌ Result failed 1/3 times: To start chrome in Debug mode, you need to close all existing Chrome instances and try again otherwise we can not connect to the instance. INFO [agent]
@alexisCode50 I also encountered this problem, my application scenario is localized Ollama, the model is deepseek. I found that using vision with uncheck off worked just fine. So I guess because the model doesn't support vision, or when you use something like a model that doesn't support vision, Error will be displayed。Hope this helps you...
Checking the "Use Own Browser" box in web ui is not working, no errors but when I enter my profile path in CHROME_USER_DATA, it gives following error:
ERROR [browser] Failed to initialize Playwright browser: BrowserType.launch: Pass user_data_dir parameter to 'browser_type.launch_persistent_context(user_data_dir, **kwargs)' instead of specifying '--user-data-dir' argument
I think it's something to do with deprecation stuff, I tried to tinker a bit, managed to get rid of the error but the browser launched was still not the profile I want it to load from CHROME_USER_DATA.
I'm having this same issue. I've tried disabling vision, trying using my own browser and not using my own browser. I also tried python 3.11 and 3.13, as well as the .venv method and the docker method. All led to the same issue. I was testing this on Windows 11. I'm going to try on my linux vm to see if the issue is the same.
I'm having this same issue. I've tried disabling vision, trying using my own browser and not using my own browser. I also tried python 3.11 and 3.13, as well as the .venv method and the docker method. All led to the same issue. I was testing this on Windows 11. I'm going to try on my linux vm to see if the issue is the same.
Same problem on Ubuntu. Both times I filled in the "CHROME_PATH" setting, but not "CHROME_USER_DATA"
Same issue here on Windows
facing same issue
已收到,谢谢!