Gets rollback automatically while creating infrastructure

[viku85]

When using api-gateway-dev-portal it executes for a while and then gets rollback automatically.

Can someone help out with the reason or any guidelines to resolve it?

[ghost]

Could you provide more details about the error it's encountering? You should be able to look in the stack and see what resource failed as well as what error caused it to roll back the deployment.

[praveenrengarajan]

@amazon-meaisiah facing the same issue

Note: am using my personal AWS account (administrator)

[ghost]

@praveenrengarajan Could you show me the precise error that first occurred before the rest of the deployment was rolled back?

I see the following were not created:

• DevPortalSiteS3Bucket
• CustomersTable
• CognitoUserPoolClientSettingsBackingFnRole
• CatalogUpdaterLambdaExecutionRole
• CloudFrontOriginAccessIdentity
• AssetUploaderLambdaExecutionRole
• LambdaEdgeFunctionRole
• CognitoUserPoolDomainBackingFnRole
• CloudFrontEdgeReplicatorRple
• LambdaCommonLayerb90953e8ec
• PreLoginAccountsTable
• CognitoPreSignUpTriggerExecutionRole

I see the following were cancelled:

• DevPortalSiteS3Bucket
• CognitoUserPoolClientSettingsBackingFnRole
• CatalogUpdaterLambdaExecutionRole
• CloudFrontOriginAccessIdentity
• AssetUploaderLambdaExecutionRole
• LambdaEdgeFunctionRole
• CognitoUserPoolDomainBackingFnRole
• CloudFrontEdgeReplicatorRple
• LambdaCommonLayerb90953e8ec
• PreLoginAccountsTable
• CognitoPreSignUpTriggerExecutionRole

What was the failure reason for CustomersTable?

[mbreuker]

I also get an automatic rollback with this error:

[mbreuker]

Found the solution to my error in #425

[mbreuker]

@amazon-meaisiah facing the same issue

Note: am using my personal AWS account (administrator)

You need to make sure that the S3 buckets that you specify, don't already exist. The first time I tried deploying the application I used existing buckets which caused a rollback.

[Jyoti-Singh-XFT]

I am still facing the same issue even after adding the fix mentioned in #425

[ghost]

@Jyoti-Singh-XFT Have you checked that the above S3 buckets don't already exist? If so, could you give some more details on the precise error you're running into (specifically, what's causing the rollbacks)?

[praveenrengarajan]

Where to update the config in the source code that all resources for dev-portal should be in ap-southeast-2 ??

[ghost]

@praveenrengarajan What command are you running to try to deploy it? If you're using SAM, you should be able to just use sam package --region ap-southeast-2 + sam deploy --region ap-southeast-2 and call it a day. If you're using node run release per the local development instructions, #505 should unblock you once it's merged.

[praveenrengarajan]

@amazon-meaisiah will cloudfront lambda@edge deploy in ap-southeast-2 ?

[ghost]

No, it must be in us-east-1 due to CloudFront restrictions. The custom resource in the template used for creating this essentially exists only to work around this SAM issue, where they don't really have special support for Lambda@Edge beyond a very rudimentary level.

[talipini]

Hi - I am getting exactly same errors. Any help?

[asheroto]

I'm having the same error, followed this verbatim, clicked on https://serverlessrepo.aws.amazon.com/applications/arn:aws:serverlessrepo:us-east-1:563878140293:applications~api-gateway-dev-portal (didn't use SAM), except that I got an error that says CognitoDomainNameOrPrefix is required so I set it.

Every time I deploy it it just rolls back.