Auth resource stopped working with notifications/analytics. Stuck on `resource auth/...`

I’ve run into this as well for the past few weeks when deploying from our dev branch to our test branch for deployment to our test env.

Stuck at the exact same spot, Build resource auth/… Then times out after 30 minutes.

Using the CLI amplify push works fine.

I thought I had it sorted last week by doing the following…

1. amplify notifications remove (removed our APNS setup)
2. amplify analytics remove
3. amplify push
4. amplify notifications remove (just for good measure to remove the pinpoint application if it’s still there)
5. check the web console and ensure that there is no application listed for the env (in our case test) in the Pinpoint service.
6. amplify analytics add
7. amplify push (important step to get the Pinpoint application setup in AWS).
8. amplify notifications add

At this point checking Pinpoint on the web console should show only 1 entry for the environment (if step 7 wasn’t done there would end up being two).

After I did this, things were working, last week. However when I did a merge to our test branch tonight, the problem came back. Steps I took to merge :

1. git checkout test
2. amplify env checkout test
3. git merge main (resolve conflicts and always choose the settings that were in the test branch before the merge for the amplify json files that come up as conflicts).
4. amplify env checkout test
5. git commit
6. git push

This kicks off the CI/CD but somehow the Pinpoint bug is back. I confirmed that amplify/backend/analytics/myapp/parameters.json is the same as what it was when it last worked.

I just tried my fix from above and unfortunately it is not working this time.

Amplify cli version 10.3.0

Looking for help here as well.

_Originally posted by @Etep15 in https://github.com/aws-amplify/amplify-hosting/issues/2541#issuecomment-1284931327_

[ykethan]

Hey @Etep15, thank you for reaching. The Amplify CLI team are investigating and are working on a resolution for this behaviour. We do not have a timeline on when the fix will be released. The Amplify CLI release page will be updated with the information. Please refer to https://github.com/aws-amplify/amplify-cli/releases for any updates.

[gspeicher]

It also seems that it is always asking for the notification config now after doing an amplify push in an environment (such as test) and then doing git checkout and then an amplify env checkout to my dev branch. Which is getting to be a bit burdensome. This appears to have happened with the latest release of amplify cli 10.3.0.

This part is happening to me but have not seen it mess with the actual deployed resources (so far at least). Amplify CLI 10.3.0 will not remember the APNs/FCM/Email settings and re-prompts for them every time I do amplify env checkout or amplify pull. This started happening immediately upon upgrade to 10.3.0 and did not require a push to trigger the behavior.