aws_backend_setup.md

We want to:

• upload our server build to AWS,
• configure a fleet for it and
• configure user pool, lambdas and rest_api to handle the login and session management

Steps

1. Using the IAM Management Console, to create a profile to give to the aws_backend.py script.

   {
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "VisualEditor0",
            "Effect": "Allow",
            "Action": [
                "gamelift:*",
                "apigateway:*",
                "cognito-idp:*",
                "iam:GetRole",
                "iam:PassRole",
                "iam:PutRolePolicy",
                "iam:DeleteRolePolicy",
                "iam:CreateRole",
                "iam:DeleteRole",
                "lambda:*"
            ],
            "Resource": "*"
        }
     ]
   }
   

2. I use a user called sean_gl, set permissions in IAM Management console and used aws configure import --csv file://new_user_credentials.csv to add the credentials for it on my PC.

3. Tweak command line aws_backend.py to match your project

   1. Run aws_backend.py with no parameters. This will display the default settings. Tweak the following:

      Setting	Purpose	Example
      prefix	represents a namespace for all of the other params	potato
      build_root	where to find the UE server build you want to push	E:\unreal_projects\ue5_gamelift_plugin_test\MyProject\ServerBuild\WindowsServer
      fleet_launch_path	the server exe to run	C:/game/MyProject/Binaries/Win64/MyProjectServer.exe
      profile	the iam profile to use	dave
      region	the aws region to use	us-west-2

      By specifying them on the command line. e.g.

      aws_backend.py --prefix=potato --build_root=E:\unreal_projects\ue5_gamelift_plugin_test\MyProject\ServerBuild\WindowsServer -- fleet_launch_path=C:/game/MyProject/Binaries/Win64/MyProjectServer.exe --profile=dave --region=us-west-2

4. Upload your build to GameLift

   1. Using the aws_backend.py command line you configured above add the command 'create build'. e.g.
   2. aws_backend.py --prefix=potato --build_root=E:\unreal_projects\ue5_gamelift_plugin_test\MyProject\ServerBuild\WindowsServer -- fleet_launch_path=C:/game/MyProject/Binaries/Win64/MyProjectServer.exe --profile=dave --region=us-west-2 create build
   3. Use the gamelift dashboard to verify your build has been uploaded

5. Start your fleet

   1. Using the aws_backend.py command line from above but this time use create fleet
   2. Use the gamelift dashboard to verify your fleet was created:
   3. If you see the error "botocore.errorfactory.LimitExceededException: An error occurred (LimitExceededException) when calling the CreateFleet operation: Current limit of fleets of 1 have been reached.", try specifying a different ec2 instance type. e.g. --fleet_ec2_instance_type=c5a.large

6. Initialize the User Pool

   1. Using the aws_backend.py command line from above but this time use create user_pool
   2. Use the Amazon Cognito dashboard to verify your user pool was created:

7. Install Lambdas

   1. Using the aws_backend.py command line from above but this time use create lambdas
   2. Use the Lambda dashboard to verify they were created

8. Install Rest API

   1. Using the aws_backend.py command line from above but this time use create rest_api
   2. The output will provide you with the invoke_url and instructions on how to test it using curl:
   3. Make note of this invoke_url
   4. At this point the backend is all ready to go.