Deployment fails "site: Access Denied"

[JanIHC]

Any idea why I'm getting a "site: Access Denied" error during the deployment to our environment?

PS C:\Repositories\fullstack-app> serverless deploy

serverless ⚡framework
Action: "deploy" - Stage: "dev" - Org: "janasmuth" - App: "fullstack" - Name: "fullstack-app"

database: 
  name:    database-dev
  arn:     arn:aws:dynamodb:ap-southeast-2:155608619720:table/database-dev
  region:  ap-southeast-2
  indexes: 
    gsi1: 
      name: gsi1
      arn:  arn:aws:dynamodb:ap-southeast-2:155608619720:table/database-dev/index/gsi1

site: Access Denied

permissions: 
  name: permissions-dev
api: 
  url: https://mmvbkodzkf.execute-api.us-east-1.amazonaws.com
  api: 
    openapi: 3.0.3
    paths: 
      /users/register: (max depth reached)
      /test/:          (max depth reached)
      /users/login:    (max depth reached)
      /user:           (max depth reached)
    info: 
      version: 0.0.1

21s » Serverless » Errors: "deploy" ran for 3 apps successfully. 1 failed.

PS C:\Repositories\fullstack-app>
PS C:\Repositories\fullstack-app> serverless deploy

serverless ⚡framework
Action: "deploy" - Stage: "dev" - Org: "abc" - App: "fullstack" - Name: "fullstack-app"

database: 
  name:    database-dev
  arn:     arn:aws:dynamodb:ap-southeast-2:12345:table/database-dev
  region:  ap-southeast-2
  indexes: 
    gsi1: 
      name: gsi1
      arn:  arn:aws:dynamodb:ap-southeast-2:12345:table/database-dev/index/gsi1

site: Access Denied

permissions: 
  name: permissions-dev
  arn:  arn:aws:iam::12345:role/permissions-dev

api: 
  url: https://abcd.execute-api.ap-southeast-2.amazonaws.com
  api: 
    openapi: 3.0.3
    paths: 
      /users/register: (max depth reached)
      /test/:          (max depth reached)
      /users/login:    (max depth reached)
      /user:           (max depth reached)
    info: 
      version: 0.0.1

24s » Serverless » Errors: "deploy" ran for 3 apps successfully. 1 failed.

[eahefnawy]

Hmmm that's odd 🤔 ... Does your AWS credentials have admin access? Could you try deploying the website independently by running the following command:

cd site
sls deploy

[juampick]

I am having the same type of issues. When deploying first time or directly from root seems that is working ok. But inside site is throwing permission denied issues.

[JanIHC]

@eahefnawy Same result if I deploy from the site folder and the AWS credentials have admin access.

PS C:\Repositories\fullstack-app\site> serverless deploy

serverless ⚡framework
Action: "deploy" - Stage: "dev" - Org: "fdsfd" - App: "fullstack" - Name: "site"

30s » Serverless » Error: Access Denied

[eahefnawy]

Org: "janasmuth"

Org: "abc"

Org: "fdsfd"

Strange, why do you have a different org everytime? 🤔

@JanIHC Could you let me know your org name and your username in the Serverless Dashboard? I'll try to look this up in our internal logs.

[JanIHC]

@eahefnawy my org name on serverless.com is janasmuth please ignore the other ones.

Could it be an issue that I'm trying to deploy to 'ap-southeast-2'?

[JanIHC]

I ran serverless info --debug but still not sure what's wrong.

Will serverless components try to activate Public access on the bucket? We have deactivated public access on account level.

PS C:\Repositories\fullstack-app\site> serverless info --debug
Fetching App Info...

serverless ⚡framework
Action: "info" - Stage: "dev" - Org: "janasmuth" - App: "fullstack" - Name: "site"
Last Action:  deploy (a few seconds ago)
Deployments:  16
Status:       error
AccessDenied: Access Denied
    at Request.extractError (/var/runtime/node_modules/aws-sdk/lib/services/s3.js:816:35)
    at Request.callListeners (/var/runtime/node_modules/aws-sdk/lib/sequential_executor.js:106:20)
    at Request.emit (/var/runtime/node_modules/aws-sdk/lib/sequential_executor.js:78:10)
    at Request.emit (/var/runtime/node_modules/aws-sdk/lib/request.js:683:14)
    at Request.transition (/var/runtime/node_modules/aws-sdk/lib/request.js:22:10)
    at AcceptorStateMachine.runTo (/var/runtime/node_modules/aws-sdk/lib/state_machine.js:14:12)
    at /var/runtime/node_modules/aws-sdk/lib/state_machine.js:26:10
    at Request.<anonymous> (/var/runtime/node_modules/aws-sdk/lib/request.js:38:9)
    at Request.<anonymous> (/var/runtime/node_modules/aws-sdk/lib/request.js:685:12)
    at Request.callListeners (/var/runtime/node_modules/aws-sdk/lib/sequential_executor.js:116:18)
State:
region:     ap-southeast-2
bucketName: website-k0w9hna
bucketUrl:  http://website-k0w9hna.s3-website-ap-southeast-2.amazonaws.com
Outputs:
Serverless » Success

Full details: https://app.serverless.com/janasmuth/apps/fullstack/site/dev
Serverless » App info fetched

[JanIHC]

@eahefnawy do you have any other ideas what could be the issue or what I could try?

[tapegram]

Wanted to bump this as I'm seeing the same issue! Are there some aws permissions I am missing? My cli user has the AdministratorAccess policy so I'm not sure what is left to try adding.

[tapegram]

I was able to find a workaround for this by giving the site’s serverless.yml a bucketName with a . in it, since this forces it to skip acceleration. It appears this is happening because of the acceleration step, but I do not know how to fix that.

[rehrumesh]

Have the same issue

[kevin4dhd]

any solution?

[kevin4dhd]

I found the solution, I had to send an email since I did not have cloudfront activated and wait 2 days for it to be activated, I tried again and everything works correctly

[mattreddy3]

Hi all, one thing that got me past this point was adding S3, all permissions (hard to tell which one was needed from the error message). The other IAM errors were well-communicated but this was opaque. Hope it helps!

[nicmeriano]

I'm having the same issue but only for one of my two environments (Dev & Prod). Whenever I try to deploy to Prod, the site component fails with an Access denied error, while everything works fine for Dev. These are associated with separate AWS accounts and I linked each stage in the Serverless console with a dedicated provider pointing to the associated account. I then deploy with sls deoploy --stage prod --aws-profile prod and while I see all the resources being created in the right account (API, permissions, Dynamo table), S3 fails.

If I try to change the S3 bucket name (following @tapegram 's advice) within serverless.yml I get the following error:

site: Changing the bucket name from site-prod to site.prod will remove your infrastructure.  Please remove it manually, change the bucket name, then re-deploy.

Both IAM users have admin access.