EAS Update Gateway Timeout


The past few weeks we’ve noticed our GitHub Action CD EAS Update deployments began to fail very frequently with Gateway Timeout errors. This doesn’t occur 100% of the time, but right now it happens for around 95% of deployments (with patches of succeeding builds). Our project itself has been migrated to SDK 45 and the issue still occurs.

The execution of the workflow look like this:

Note: eas-cli version is 0.52.0

The workflow YAML looks like this:

    name: "Deploy Mobile App"
    needs: prepare_sentry_release
    runs-on: ubuntu-latest
      - name: "Checkout repository"
        uses: actions/checkout@v1

      - uses: ./.github/actions/yarn

      - name: Setup Expo
        uses: expo/expo-github-action@v6
          expo-version: latest
          eas-version: latest
          token: ${{ secrets.EXPO_TOKEN }}
          expo-cache: true
          eas-cache: true

      - name: "Build Slate Editor Harness and Workspace TypeScript"
        run: yarn workspace @attio/mobile-app eas-build-post-install

      - name: Publish EAS update
        working-directory: packages/runtimes/mobile-app
        run: eas update --branch production --message "${{ github.event.head_commit.message }}"
          EXPO_SLUG: mobile-app
          NODE_OPTIONS: --max_old_space_size=4096
          SENTRY_AUTH_TOKEN: ${{ secrets.SENTRY_AUTH_TOKEN }}
          SENTRY_RELEASE: ${{ github.sha }}

Any guidance on this would be great - we’re not sure what changed on our end.

All the best,

Thank you for your report. We will look into this.

1 Like

Could you run expo export --experimental-bundle? This should create a dist folder in your project. If you can zip that folder up and share with us (you can send me the file or download link in a private message if you want), then we can take a look and see if anything in the bundle is causing this.

For what it’s worth, I can’t seem to replicate the timeout locally - only seems to happen on CI.
I have the bundle zipped up locally, but can’t seem to find how to message you :sweat_smile:

@kbrandwijk I think you would have to message @bradenattio first so that he can reply :slight_smile:

Yup, did done that.


Hello :wave:
Wondering if you have any updates regarding this?
Over the past ~30 deployments, we’ve had 1 successful mobile EAS Update deployment (5 days ago). It’s causing quite an interruption to our deployment pipeline :frowning:

Yes! There seems to be an issue in our API related to the large number of updates you have on a single branch. We have a fix for that in development, I hope we’ll be able to deploy this soon, but I’ll keep you posted once I have an ETA on it.


Hey, just writing to catch-up on the progress of this. Our CD pipeline has been blocked for ~1 month, which is quite disruptive :frowning: Perhaps there is a quick temporary workaround? Maybe we can release a new native build with production-new as the branch?

Hi, I’m sorry for the delay here, the original issue isn’t fully resolved yet.

And yes, pushing new updates to a new branch should work fine. You don’t need to run a new native build for that, you can edit your existing channel to point at the new branch (one of the benefits of EAS Update).

@bradenattio We have released an update to eas-cli, if you install 0.53.1, then you shouldn’t have the issue anymore on your old branch either. Can you try that out for me and let me know? Thanks!

1 Like

Awesome - that seems to have resolved our issues! Thanks as always :slight_smile:

1 Like

This topic was automatically closed 20 days after the last reply. New replies are no longer allowed.