I have an Amplify app, but when I deploy a new version, the build is timing out. When I check CloudFormation, one of the nested stacks is in the status UPDATE_ROLLBACK_IN_PROGRESS. Any ideas why this error is occurring?
#UPDATE_ROLLBACK_IN_PROGRESS stuck status
south moss
gaunt wave
👋 this typically pops up when a deployment fails or something like a container failed to start -- this would cause the container to loop until CFN timed out. If you look at the nested stack in question and look through the events for a failure, does it give a decent reason?
south moss
Yes I have a custom resource that failed to deploy. when I click in Detect root cause I see the reason (I added the xxxx to clean the names): Embedded stack arn
cloudformation:us-east-1xxxstack/amplify-xxxxxxxxx-customcustomResource/xxxxx was not successfully updated. Currently in UPDATE_ROLLBACK_IN_PROGRESS with reason: The following resource(s) failed to update: [AllowBucketNotificationsTolakeFileTriggerFunctionReferences, xxxxxxNotificationsxxxxx, xxxxxxxxxPolicy19C94EDF]. When I go to the nested stack the last event is in UPDATE_ROLLBACK_COMPLETE_CLEANUP_IN_PROGRESS . The failed reason of the nested stack is: Received response status [FAILED] from custom resource. Message returned: Error: An error occurred (InvalidArgument) when calling the PutBucketNotificationConfiguration operation: Unable to validate the following destination configurations. See the details in CloudWatch Log Stream: 2024/02/07/[$LATEST]xxxx . Any ideas how can I solve this?
When I try to re deploy the Amplify App, it fails after 3 min with the reason: ["Index: 0 State: {"deploy":"deploying"} Message: Stack:arn:aws:cloudformation:us-east-1❌xxxxxxxxxxxxxxxstack/amplify-xxxxxxxx is in UPDATE_COMPLETE_CLEANUP_IN_PROGRESS state and can not be updated."]
gaunt wave
Error: An error occurred (InvalidArgument) when calling the PutBucketNotificationConfiguration operation: Unable to validate the following destination configurations
this is an interesting error. What does the custom resource do? is this using a Function to ingest some data and send to S3? Are you using S3 from amplify add storage?
south moss
I just solve the error. and the cause is weird, let me explain. I have 2 s3 bukcets, the first one (added with amplify add storage) and I have another s3 bucket added as a custom resourse. (i have 2 buckets becasuse I wanted to separete business logic permissions, etc).
I have a Imabda function (added as amplify add function). In the custom resource I have defined the s3 bucket and a trigger (bucket.addEventNotification(s3.EventType.OBJECT_CREATED, new notifications.LambdaDestination(importedLambda))) with the lambda function imported with the AmplifyDependentResourcesAttributes
Seems like when the Stack update fails, I assume maybe because the lambda function does not exist yet when the notification of the custom resourse is being created,
If I delete the bucket.addEventNotification the stack update works without error. And in the next stack update I can add the addEventNotification and it works without error (because the lambda function is already created in the previous push)
The question is, what is the order of the stack resourses when they are being created? is this problem related because I am combining a lambda function created with the amplify add function and a s3 bucket created as a custom resource?