karma-chrome-launcher: Having problems running karma with jenkins in chrome

I was running into the same issue with both e2e protractor tests and Karma unit tests. Windows jenkins specifically (running manually in cmd was working fine).

The (temp) fix was the same for both:

In karma.conf:

    browsers: ['ChromeNoSandbox'],

    customLaunchers: {
      ChromeNoSandbox: {
        base: 'Chrome',
        flags: ['--no-sandbox']
      }
    }

In protractor.conf

    browserName: 'chrome',
    chromeOptions: {
      args: ['--no-sandbox']
    }
  }

“karma-chrome-launcher”: “^1.0.1”, and “protractor”: “4.0.0”,

Same problem for me with the following setup:

browsers: [
  'ChromeNoSandbox'
],

customLaunchers: {
  ChromeNoSandbox: {
    base: 'Chrome',
    flags: ['--no-sandbox']
  }
},

karma-chrome-launcher: 2.0.0 karma: 1.3.0 chrome: 53.0.2785.116 os: CentOS 7

@jotka Check this comment I gave on a related issue : https://github.com/karma-runner/karma/issues/1206#issuecomment-337599491

I’m using the same versions as yours, and I can run everything without any problem now.

I am attempting to run in bamboo:

karma-chrome-launcher: 2.1.1
karma: 1.7.0
os: RHEL 6.9 (using xvfb-run to run headless)
chrome: Chromium 58.0.3029.96 (using export CHROME_BIN="chromium-browser")

Same problem:

01 06 2017 18:28:34.850:INFO [karma]: Karma v1.7.0 server started at http://0.0.0.0:9876/
01 06 2017 18:28:34.853:INFO [launcher]: Launching browser ChromeNoSandbox with unlimited concurrency
01 06 2017 18:28:35.039:INFO [launcher]: Starting browser Chrome
01 06 2017 18:29:35.090:WARN [launcher]: Chrome have not captured in 60000 ms, killing.
01 06 2017 18:29:35.578:INFO [launcher]: Trying to start Chrome again (1/2).

I had the exact same issue with our jenkins and ChromeHeadless.

This solved it for me: set a NO_PROXY env variable : export NO_PROXY=“localhost, 0.0.0.0/4201, 0.0.0.0/9876”

Source: https://github.com/karma-runner/karma/issues/1206#issuecomment-337599491

hi dear all. Has this been somehow fixed? I have the same on the lastest Chrome / Debian. “Chrome have not captured in 60000 ms, killing.” when running a job on Jenkins through docker container.

“karma”: “1.7.1”, “karma-chrome-launcher”: “2.2.0”,

I’m having this as well. It does not capture on first attempt, but does in second attept. This makes the Jenkins CI cycle be very slow. Any fix for this? (Same error on both Chrome and ChromeHeadless)

“karma”: “1.7.0”, “karma-chrome-launcher”: “2.1.1”, Chrome 60.0.3112.90 (Official Build) (64-bit)

Hello, I’m running into this as well now… after upgrading packages:

-    "karma": "~1.5.0",
-    "karma-chrome-launcher": "~2.0.0",
+    "karma": "~1.7.1",
+    "karma-chrome-launcher": "~2.2.0",

Chrome 60 is installed on Windows 7 64-bit. Tried the fix mentioned in this ticket, but it still cannot connect with --no-sandbox.

Did anyone have a work around for this?

Was experiencing this problem with Chrome 52.0.x Karma-Chrome-Launcher 1.0.1 and 2.0.0

Chrome update as of this morning to 53.0.2785 and seems to have resolved it. Anyone else confirm?

@agccheung Yep. Can confirm that it is now working for us in Chrome 53.0.2785

We tried edgahan’s temp fix but it did not work for us. The issues started around upgrading to karma-chrome-launcher: 1.0.1 and Chrome 51 upgrading to 52. It’s unclear what’s the issue here since our Jenkins machine spits out:

12:49:29      [exec] [1] START:
12:49:29      [exec] [0] npm run lint exited with code 0
12:49:29      [exec] [1] 01 08 2016 12:49:22.360:INFO [karma]: Karma v1.1.1 server started at http://localhost:9876/
12:49:29      [exec] [1] 01 08 2016 12:49:22.362:INFO [launcher]: Launching browsers ChromeNoSandbox, ElectronNoSandbox with unlimited concurrency
12:49:29      [exec] [1] 01 08 2016 12:49:22.375:INFO [launcher]: Starting browser Chrome
12:49:29      [exec] [1] 01 08 2016 12:49:22.392:INFO [launcher]: Starting browser electron
12:49:29      [exec] [1] 01 08 2016 12:49:25.979:INFO [Chrome 52.0.2743 (Windows 8.1 0.0.0)]: Connected on socket /#80VNe5WGo53GZH1OAAAA with id 40667921
12:49:29      [exec] [1] Chrome 52.0.2743 (Windows 8.1 0.0.0) ERROR
12:49:29      [exec] [1]   Script error.
12:49:29      [exec] [1] 01 08 2016 12:49:27.506:INFO [Chrome 47.0.2526 (Windows 8.1 0.0.0)]: Connected on socket /#q_QvLGea2t_M6n-fAAAB with id 49218709
12:49:29      [exec] [1] Chrome 47.0.2526 (Windows 8.1 0.0.0) ERROR
12:49:29      [exec] [1]   Script error.
12:49:29      [exec] [1] 
12:49:29      [exec] [1] Finished in 2.826 secs / 0 secs
12:49:29      [exec] [1] 
12:49:29      [exec] [1] 
12:49:29      [exec] [1] SUMMARY:
12:49:29      [exec] [1] √ 0 tests completed
12:49:29      [exec] [1] babel-node node_modules/karma/bin/karma start exited with code 1
12:49:29      [exec] 

Next up, will be checking if reverting to older launcher version helps. UPDATE - no fix from reverting karma-launcher versions - likely this is related to: https://bugs.chromium.org/p/chromium/issues/detail?id=615396

Running into the same issue.

Without any changes in any JS modules, this appears to stop working after Chrome upgraded from 51.0.2704.103 to 52.0.2743.82.

Some time ago our e2e tests using a selenium server on Windows started to fail due to this issue. We have yet to validate if that still happens, but it might be related?

We’re not using Jenkins here (Microsoft Visual Studio Team Services) but a similar setup I guess. The build agent runs as a Windows Service in session 0, spawns a build process (Gulp in our case) which starts Chrome using this plugin.