amplify-cli: Ampilfy push or build gives me "Invalid AttributeDataType input" error after adding analytics to existing project auth that was building fine

anyone of you with this problem even when the required attributes are in sync? I have a custom attribute added but it worked for months without problems.

I am having the same issue

anyone of you with this problem even when the required attributes are in sync? I have a custom attribute added but it worked for months without problems.

I have fixed my issue by following these steps:

1. check the Required attributes in User Pools

2. look into cli-inputs.json file “requiredAttributes”: [ “email”, “family_name”, “given_name” ]

3. remove both “email” and “family_name”

4. run “amplify push”

5. redeploy in the console again

6. done!

Customers in #9532 are reporting downgrade to CLI v7.6.8 works around this issue

@johnpc I’m facing the same issue and it can’t be resolved after downgrading the amplify CLI to v7.6.8 There are some custom:attributes were added manually to the Cognito. I can run amplify push couple days ago without any problem.

UPDATE_FAILED UserPool AWS::Cognito::UserPool  Fri Feb 04 2022 00:02:35 GMT+1100 (Australian Eastern Daylight Time) Invalid AttributeDataType input, consider using the provided AttributeDataType enum. (Service: AWSCognitoIdentityProviderService; Status Code: 400; Error Code: InvalidParameterException; Request ID: 5766d0a9-0141-49da-ae1f-54d5bd8e12f8; Proxy: null)

{
  "version": "1",
  "cognitoConfig": {
    "resourceNameTruncated": "amplifb7e0ab32",
    "userPoolName": "fortifyfox",
    "autoVerifiedAttributes": [
      "email"
    ],
    "mfaConfiguration": "OFF",
    "mfaTypes": [
      "SMS Text Message"
    ],
    "smsAuthenticationMessage": "Your authentication code is {####}",
    "smsVerificationMessage": "Your verification code is {####}",
    "emailVerificationSubject": "Your verification code",
    "emailVerificationMessage": "Your verification code is {####}",
    "defaultPasswordPolicy": false,
    "passwordPolicyMinLength": 8,
    "passwordPolicyCharacters": [],
    "requiredAttributes": [
      "email",
      "family_name",
      "given_name"
    ],
    "aliasAttributes": [],
    "userpoolClientGenerateSecret": false,
    "userpoolClientRefreshTokenValidity": 30,
    "userpoolClientWriteAttributes": [
      "email"
    ],
    "userpoolClientReadAttributes": [
      "email"
    ],
    "userpoolClientLambdaRole": "fortifb7e0ab32_userpoolclient_lambda_role",
    "userpoolClientSetAttributes": false,
    "authSelections": "userPoolOnly",
    "resourceName": "fortifyfox",
    "serviceName": "Cognito",
    "useDefault": "manual",
    "sharedId": "b7e0ab32",
    "userPoolGroupList": [],
    "userPoolGroups": false,
    "usernameAttributes": [
      "email"
    ],
    "usernameCaseSensitive": false,
    "adminQueries": false,
    "hostedUI": true,
    "triggers": {
      "CustomMessage": [
        "custom"
      ],
      "PostConfirmation": [
        "add-to-group",
        "custom"
      ],
      "PreSignup": [
        "custom"
      ]
    },
    "authRoleArn": {
      "Fn::GetAtt": [
        "AuthRole",
        "Arn"
      ]
    },
    "unauthRoleArn": {
      "Fn::GetAtt": [
        "UnauthRole",
        "Arn"
      ]
    },
    "breakCircularDependency": true,
    "useEnabledMfas": false,
    "dependsOn": [
      {
        "category": "function",
        "resourceName": "fortifyfoxCustomMessage",
        "triggerProvider": "Cognito",
        "attributes": [
          "Arn",
          "Name"
        ]
      },
      {
        "category": "function",
        "resourceName": "fortifyfoxPostConfirmation",
        "triggerProvider": "Cognito",
        "attributes": [
          "Arn",
          "Name"
        ]
      },
      {
        "category": "function",
        "resourceName": "fortifyfoxPreSignup",
        "triggerProvider": "Cognito",
        "attributes": [
          "Arn",
          "Name"
        ]
      }
    ],
    "permissions": [
      "{\n  \"policyName\": \"AddToGroupCognito\",\n  \"trigger\": \"PostConfirmation\",\n  \"effect\": \"Allow\",\n  \"actions\": [\n    \"cognito-idp:AdminAddUserToGroup\",\n    \"cognito-idp:GetGroup\",\n    \"cognito-idp:CreateGroup\"\n  ],\n  \"resource\": {\n    \"paramType\": \"!GetAtt\",\n    \"keys\": [\n      \"UserPool\",\n      \"Arn\"\n    ]\n  }\n}"
    ],
    "authTriggerConnections": "[\n  {\n    \"triggerType\": \"CustomMessage\",\n    \"lambdaFunctionName\": \"fortifyfoxCustomMessage\"\n  },\n  {\n    \"triggerType\": \"PostConfirmation\",\n    \"lambdaFunctionName\": \"fortifyfoxPostConfirmation\"\n  },\n  {\n    \"triggerType\": \"PreSignUp\",\n    \"lambdaFunctionName\": \"fortifyfoxPreSignup\"\n  }\n]",
    "parentStack": {
      "Ref": "AWS::StackId"
    },
    "hostedUIDomainName": "dev-googlesignin",
    "authProvidersUserPool": [
      "Google"
    ],
    "hostedUIProviderMeta": "[{\"ProviderName\":\"Google\",\"authorize_scopes\":\"openid email profile\",\"AttributeMapping\":{\"email\":\"email\",\"family_name\":\"family_name\",\"given_name\":\"given_name\",\"username\":\"sub\"}}]",
    "oAuthMetadata": "{\"AllowedOAuthFlows\":[\"implicit\"],\"AllowedOAuthScopes\":[\"phone\",\"email\",\"openid\",\"profile\",\"aws.cognito.signin.user.admin\"],\"CallbackURLs\":[\"http://localhost:8878/\",\"https://dev.fortifyfox.com/\"],\"LogoutURLs\":[\"http://localhost:8878/\",\"https://dev.fortifyfox.com/\"]}",
    "authProviders": []
  }
}

I was finally able to get amplify push to work. It appears that it was the same problem with requiredAttributes, but it was presenting itself differently.

Required Attributes in the Cognito Console showed both name and email:

(NOTE: replaced my app id with ‘myapp1234’ in the paths below)

amplify/backend/auth/myapp1234/cli-inputs.json had:

“requiredAttributes”: [ “email”, “name” ],

amplify/backend/auth/myapp1234/build/myapp1234-cloudformation-template.json had:

“Schema”: [ { “Mutable”: true, “Name”: “email”, “Required”: true }, { “Mutable”: true, “Name”: “name”, “Required”: true }

amplify/#current-cloud-backend/auth/myapp1234/cli-inputs.json had:

“requiredAttributes”: [ “email”, “name” ],

amplify/#current-cloud-backend/auth/myapp1234/build/myapp1234-cloudformation-template.json had:

“Schema”: [ { “Mutable”: true, “Name”: “name”, “Required”: true } ],

So the only incorrect file was the cloudformation-template file in the #current-cloud-backend directory. Everything else appeared correct.

I removed “email” from the amplify/backend/auth/myapp1234/cli-inputs.json file as @johnpc suggested. I then did an amplify push and everything worked. All the above files now show only the name attribute.

Strangely, the Cognito console still shows both “name” and “email”. It does not appear to be causing a problem for me though.

This obviously does not solve the root issue that caused the sync drift, but it got me up and running again after several days of frustration and downtime.

Hopefully this helps everyone else that has this issue.

Hi, I have fixed my issue => Pushing only my auth again have fixed the issue my steps:

1. Modify something inside cli-inputs.json of backend auth (positon of attributes etc…) (to be able to update the auth with amplify push)
2. Amplify push After the Amplify update, I was able to push and build again… It’s works for me but i can’t explain what was the issue.

Hi team @josefaidt @Straubulous any updates on this issue, my team is blocked for almost two weeks now with this error. I have sent my project my project identification number: 8917f531808b98396e5b186f78b710ba. I have read through the whole discussions and none of the suggestions worked. Thanks

I am also still getting it on 10.4.0. It seems not much is happening on this issue.

Initial fix addressed part of this. Additional documentation to describe high level approach for migration of user pool.

@josefaidt sent a message on discord 👍🏻

Hey @Shomari and folks in the thread 👋 are you still experiencing this issue? There’s quite a few comments here describing the affected workflow and a fix has since gone out for this issue, though I wanted to be sure this was taken care of 🙂

@voyagebagage I believe I may have been in a similar situation as you and I was able to work through it. The following worked for me:

1. Revert to the amplify-cli version that was being used before trying to update (in my case 7.6.8)
2. Edit the cli-inputs.json to remove the problem requiredAttributes (the ones not in the cloudformation template)
3. amplify push
4. Update amplify-cli (in my case to 7.6.15)
5. amplify push --force

@voyagebagage , check your filesystem. I’m using VSCode for this project, and it hides the #current-cloud-backend folder in the explorer inside VSCode.

@jladdison 's solution above helped me but it’s certainly a strange error. My Cloudformation template all matched except for the one that was being pulled from remote. What’s stranger is that I hadn’t actually done a push on the project in quite some time so I can’t really pinpoint when the requiredAttributes property was updated. In short, nothing had changed until I updated the CLI version. Seems the initial stages. of the push with the new CLI version modified the CF template and dropped all but one requiredAttribute?

So only “picture” is required in your deployed Cognito environment, but your desired configuration is:

"requiredAttributes": [
  "email",
  "family_name",
  "given_name",
  "picture"
],

The way this looks is, your deployment was bitten by the bug that https://github.com/aws-amplify/amplify-cli/pull/9251 attempted to fix - only one of your required attributes was set in reality. However, since requiredAttributes cannot be edited after the Cognito resource is deployed (a Cognito limitation), it is causing an error where your deployments are failing.

I think this makes sense why downgrading is working for so many people. They’re re-pushing with missing requiredAttributes when they downgrade, so the push matches what’s in the cloud. Another “fix” to unblock deploying new changes would be to update requiredAttributes in your cli-inputs.json to match what’s in the cloud (aka, change it so only “picture” is required")

The real long term fix would be to remove your Cognito configuration and replace it with a new one. Of course, you do not want to lose existing user data, so you’d have to migrate your users to a new, properly configured Cognito resource and then import the new auth resource to your app.

Customers in https://github.com/aws-amplify/amplify-cli/issues/9532 are reporting downgrade to CLI v7.6.8 works around this issue

I’m getting this too!