8

I've added the google sign in to my app and tried to push the modifications, however I didn't have permissions to do that. After I got the permissions, I tried to amplify push but I get this error:

Stack:arn:aws:cloudformation:us-east-1:432853258441:stack/amplify-huruproject-local-190356/117b7360-c94d-11eb-a4cc-0e8a861a6983 is in UPDATE_ROLLBACK_FAILED state and can not be updated.

An error occurred during the push operation: Stack:arn:aws:cloudformation:us-east-1:432853258441:stack/amplify-huruproject-local-190356/117b7360-c94d-11eb-a4cc-0e8a861a6983 is in UPDATE_ROLLBACK_FAILED state and can not be updated.

Thing is I can't find this specific stack in my stacks console. How can I fix this? UPDATE: I found the stack, it was in an autogenerated nested stack. this is what I found: enter image description here

enter image description here enter image description here

I don't know how to proceed from here. should I delete the whole stack? Will I lose all my data?

confusedstudent
  • 353
  • 3
  • 11

2 Answers2

11

A little late to reply but I'll just put this here for future reference. Deleting the stack or continuing are the only two options you have when you hit the "UPDATE_ROLLBACK_STATE". But don't go deleting the stack right away.

STEP 01.

  • Try Continuing the Update. enter image description here

STEP 02.

  • If it still goes to the "UPDATE_ROLLBACK_STATE", click "Continue Update Rollback" again
  • Expand "Advanced Trouble shooting" and check the resources to skip. (You will probably have to check them all)
  • Then click "continue Update Rollback"

enter image description here

Step 03

  • After step 02, you should be at a "Update Rollback Success" state.
  • Before deploying again, try to find out the cause for the failure.
  • For me it was deploying a service with Node V10 and it wasn't supported for AWS Lambda deployment no more.
  • Fix what ever is causing the deployment issue and redeploy.

STEP 04

  • Sadly if non of the above steps worked, then you will have to delete the stack.
  • This is a nuisance especially if you're using services like API Gateway or DynamoDB streams. You'll have to update all the relevant links in your front-end afterwards.
iWiiCK
  • 361
  • 5
  • 13
2

As per the documentation, you can either delete the stack or try continuing the rollback. The simplest solution would be to just delete the stack and try again.

Here's the summary:

If your stack is stuck in the UPDATE_ROLLBACK_FAILED state after a failed update, then the only actions that you can perform on the stack are the ContinueUpdateRollback or DeleteStack operations. This is because CloudFormation requires further input from you to acknowledge that the stack is out of sync with the template that the stack is attempting to roll back to. To retry the rollback and resolve the error, you can use ContinueUpdateRollback.

Tip: To resolve the error, you might need to raise limits, change permissions, or modify other settings. For information on common errors that cause update rollback failures, see Update rollback failed.

In some cases, retrying the rollback doesn't resolve the error. In these cases, you must skip resources, and also acknowledge that these resources no longer reflect the desired state in the CloudFormation template. To skip the resources that caused the rollback failures, complete the following steps in the Resolution section.

As for the cause of the error - you can see from the CloudFormation messages that the readAttributes and writeAttributes of a UserPoolClient cannot be empty, like they are in your case.

gshpychka
  • 8,523
  • 1
  • 11
  • 31