New issue
Advanced search Search tips

Issue 905943 link

Starred by 1 user

Issue metadata

Status: Assigned
Owner:
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux
Pri: 2
Type: Bug



Sign in to add a comment

--remote-debugging-address is only supported if --headless is specified

Reported by killiank...@gmail.com, Nov 16

Issue description

UserAgent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.102 Safari/537.36

Steps to reproduce the problem:
1. Launch chrome with --remote-debugging-address 
google-chrome-stable --remote-debugging-port=9222 --remote-
debugging-address=0.0.0.0
2. Unable to connect through anything but the loopback interface, ex:
ERR - curl <ip>:9222
OK  - curl localhost:9222

What is the expected behavior?

--remote-debugging-address should function the same way whether or not --headless is specified

Public documentation does not indicate that --headless is required:
https://peter.sh/experiments/chromium-command-line-switches/

What went wrong?

Did this work before? No 

Chrome version: 70.0.3538.102  Channel: stable
OS Version: Ubuntu 18.04.1 LTS
Flash Version: 

Others have ran into this discrepancy, see:
- https://stackoverflow.com/questions/40538197/chrome-remote-debugging-from-another-machine
- https://stackoverflow.com/questions/48319240/chrome-fails-to-serve-from-debugging-port-inside-docker-container
- https://stackoverflow.com/a/39982731

Furthermore, a fix here would simplify a common remote debugging scenario:
- https://stackoverflow.com/questions/18506233/using-chromium-remote-debugging-from-external-device
- https://coderwall.com/p/kmoe2a/chrome-truly-remote-debugging
 
The justification for the original --headless CL is still relevant here as well for folks using chrome inside docker with xvfb. See https://codereview.chromium.org/1920773003

"The stated use case was to expose the debugging protocol from within a
docker container (which can't be done for a loopback address)."




Labels: Needs-Triage-M70
Owner: caseq@chromium.org
Status: Assigned (was: Unconfirmed)
We should probably remove the --remote-debugging-address from headless to avoid confusion. Tunnel should be used to make the port accessible, but we need to verify this is docker-friendly before we do that.
Assuming folks want to keep --remote-debugging-address I have a CL ready to submit when the bug is approved.

Sign in to add a comment