gitlab ci blocks during run

[chenfan0]

lighthouserc.js

module.exports = {
    ci: {
        collect: {
            url: [
                'http://127.0.0.1:8089/'
            ],
            numberOfRuns: 3,
            startServerCommand: "PORT=8089 node src/server",
            settings: {
                hostname: '127.0.0.1',
                chromeFlags: ['--no-sandbox'],
            },
        },
        upload: {
            target: 'temporary-public-storage',
        }
    }
}

.gitlab-ci.yml

lhci:
  stage: test
  image: cypress/browsers:node-18.16.1-chrome-114.0.5735.133-1-ff-114.0.2-edge-114.0.1823.51-1
  script:   
    - npm install
    - npm install -g @lhci/cli@0.13.x
    - npm run build
    - lhci autorun --config=./lighthouserc.js || echo "LHCI failed!"

[ydhnwb]

I also got this. Any solution?

[njsaunders]

Has the npm package been compromised?

$ npx lhci --version
Hello, this is AnupamAS01!

[chenfan0]

I also got this. Any solution?

no 😞

[lukaas33]

I have the same issue with the following CI config:

lighthouse:
  stage: .pre
  image: cypress/browsers:node16.17.0-chrome106
  script:
    - npm install -g @lhci/cli@0.13.x
    - lhci autorun --upload.target=temporary-public-storage --collect.settings.chromeFlags="--no-sandbox" --collect.staticDistDir="wiki/pages" --collect.numberOfRuns=1 || echo "LHCI failed!"

[chenfan0]

I found a solution. I executed this download-chrome.sh to install Chrome, and then set the environment variable LIGHTHOUSE_CHROMIUM_PATH to the location of Chrome downloaded by the script, and then it can run normally.

[ArgV04]

Any news on this? I am also facing this kind of issue

[chenfan0]

Any news on this? I am also facing this kind of issue

I found a solution. I executed this download-chrome.sh to install Chrome, and then set the environment variable LIGHTHOUSE_CHROMIUM_PATH to the location of Chrome downloaded by the script, and then it can run normally.

here

[ArgV04]

Any news on this? I am also facing this kind of issue

I found a solution. I executed this download-chrome.sh to install Chrome, and then set the environment variable LIGHTHOUSE_CHROMIUM_PATH to the location of Chrome downloaded by the script, and then it can run normally.

here

How does it work? The script already checks for LIGHTHOUSE_CHROMIUM_PATH and exit if not exist. And what the value of LIGHTHOUSE_CHROMIUM_PATH ... I tried this script partially ... meaning downloading chrome .... and execute wget ... but with this ... its still not working

[njsaunders]

Manage to "Fix" this by using Node 22 and setting the following flag:

export NODE_OPTIONS="--max-old-space-size=7168"

[njsaunders]

Turns out that was only for one pipeline. It's happening again. Have opened #1070 to figure out whats going on. Not giving up on this one!

[njsaunders]

With debugging turned on looks like it's blocking connecting to Chrome:

Mon, 02 Sep 2024 12:13:45 GMT LH:status Auditing: Document has a valid `rel=canonical`
Mon, 02 Sep 2024 12:13:45 GMT LH:status Auditing: Structured data is valid
Mon, 02 Sep 2024 12:13:45 GMT LH:status Auditing: Page didn't prevent back/forward cache restoration
Mon, 02 Sep 2024 12:13:45 GMT LH:status Generating results...
done.
Run #3...Mon, 02 Sep 2024 12:13:46 GMT LH:ChromeLauncher Found existing Chrome already running using port 41471, using that.

Just sits there waiting at that. Debugging continues.

[njsaunders]

Ok - I don't know why but this is related to the version of Chrome used. I downgraded from 128 to 127 and everything started working fine again.

[alexislefebvre]

We faced the same issue, we fixed it by using a newer version of the container, it has updates of node and Chrome, so we don't know the root cause.

This never ends:

docker run --rm -it cypress/browsers:node-20.5.0-chrome-114.0.5735.133-1-ff-114.0.2-edge-114.0.1823.51-1 \
bash -c 'yarn global add @lhci/cli@0.14.x ; lhci collect --url https://www.google.com \
--collect.settings.chromeFlags="--no-sandbox"'

This works:

docker run --rm -it cypress/browsers:node-20.17.0-chrome-128.0.6613.119-1-ff-130.0-edge-128.0.2739.63-1 \
bash -c 'yarn global add @lhci/cli@0.14.x ; lhci collect --url https://www.google.com \
--collect.settings.chromeFlags="--no-sandbox"'

[kevinprakasa]

This has also happened to me all of a sudden without any recent changes, my lighthouse CI in github actions took forever until it passed the time limit. And I managed to fix the issue.

For my case, I'm using the chrome version from puppeteer@13.5.2 package that we're using to authenticate the LHCI request. I believe that is still using older chrome version than the latest version now. I tried to update puppeteer to version 23.7.0, and suddenly, the lighthouse running process is no longer stuck anymore. Apparently, using Chrome/130.0.6723.58 that is coming from the latest puppeteer fixed the issue for me.

But I wonder is there any flag command like --debug when running the LHCI so that I can see the core problem behind all this?

[tractorcow]

Has the npm package been compromised?

$ npx lhci --version
Hello, this is AnupamAS01!

npm install -g @lhci/cli not npm install -g lhci. I too made this hilarious mistake.