Giter Site home page Giter Site logo

aws-samples / automate-saml-federation Goto Github PK

View Code? Open in Web Editor NEW
10.0 3.0 5.0 23 KB

AWS IAM Role sync with Azure Active Directory | extensible for other Directory Services

License: MIT No Attribution

Python 66.66% Shell 33.34%
aws aws-organizations saml azure-active-directory iam sso sso-authentication sso-solution lambda-functions aws-lambda

automate-saml-federation's Introduction

About:

This is a sample solution to integrate multiple AWS Organization Member accounts with a 3rd party SAML provider for SSO. It provides a Lambda function which can be scheduled to run at specific intervals by CloudWatch and which will visit all the member accounts of an AWS Organizations, read all the Roles which do trust a specific SAML IDP Provider and synchronize it with the Provider. Currently there is one single Provider Synchronizer implementation with Microsoft Azure, but the solution can be easily extended in the future.

It provides two components, packaged into a single function:

  • the main APP, which can be scheduled from CloudWatch and will collect all Roles trusting a selected SAML provider;
  • azure_sync which will synchronize the collected roles with Azure AD via the Graph (Rest Api);

The function reads all parameters from System Manager's Property Store.

Additionally there are a set of CloudFormation Templates which will:

This solution is loosely based on the original work described in this blog and these powershell scripts;

Requirements

Setup process

This guide relies on the use of CloudFormation stack set, which is a centralized way of running CloudFormation templates in multiple member accounts. First we need to prepare the organization:

  • add the AWSCloudFormationStackSetAdministrationRole and AWSCloudFormationStackSetExecutionRole to the root account;
  • and the role AWSCloudFormationStackSetExecutionRole to all of the member accounts;

Once this step is ready we need to set up our SAML provider. This template takes Azure as an SSO/Sampl Provider sample implementation. The guide to set it up can be found here.

As soon as we have the METADATA file from the SAML provider, it comes the "pièce de résistance", which is composed of:

Prepare the root account for executing Stackset instances in each and every member account

Let's prepare the root account so that to be able to execute Stack Sets:

aws cloudformation create-stack --stack-name stackset-admin-role \
--capabilities CAPABILITY_NAMED_IAM --profile root-org \
--template-body file://./cfn_templates/root_account_stackset_admin.yaml

Please note: you need to replace the "root-org" with the name of the AWS CLI profile name of your AWS Account, which represents the AWS Organization's Root Account. Check the execution status:

aws cloudformation describe-stacks --stack-name stackset-admin-role \
--profile root-org --query 'Stacks[0].StackStatus'

Prepare the member accounts

Run the following command once for every member account, using the member credentials (replace the "xxxxyyyyy" with the ID of the root account)

aws cloudformation create-stack --stack-name cross-account-access-roles \ 
--parameters ParameterKey=TrustedAccountNumber,ParameterValue=xxxxyyyyy \
--capabilities CAPABILITY_NAMED_IAM --profile member1 \
--template-body file://./cfn_templates/cross_account_access_for_invited_members.yaml

Please Note: the "member1" string needs to be changed to the name of the profile your member account;

Wait until the following command returns "CREATE_COMPLETE":

aws cloudformation describe-stacks --stack-name cross-account-access-roles \ 
--profile member1 --query 'Stacks[0].StackStatus'

In case something goes wrong you might want to check the details with:

aws cloudformation describe-stack-events \
--stack-name cross-account-access-roles \
--profile member1 \
--query 'StackEvents[*].[ResourceType,ResourceStatus,ResourceStatusReason]' \
--output text|column -s $'\t' -t

Repeat these steps with all member accounts which were invited (and not created) in your organization. Once finished, we are ready to use the CloudFormation StackSet to roll out unified configurations in our organization.

Add IAM Access Role to all accounts using CloudFormation Stacks

At this step we will create a new stack set which will deploy a restricted role, called AWS_IAM_AAD_UpdateTask_CrossAccountRole used by the role synchronizing lambda function in a later step (please not: "xxxxyyyyy" is the account number of your root account).

aws cloudformation create-stack-set --stack-set-name member-iam-access-role --profile root-org \
--parameters ParameterKey=TrustedAccountNumber,ParameterValue=xxxxyyyyy \
--capabilities CAPABILITY_NAMED_IAM \
--template-body file://./cfn_templates/member_iam_access_role.yaml

Wait until the following command returns "Active" for a stack set name "member-iam-access-role":

aws cloudformation list-stack-sets --profile root-org

or

aws cloudformation list-stack-sets --profile root-org --query "Summaries[?(@.StackSetName=='member-iam-access-role')].Status | [0]"

Create the IDP providers and a set of initial roles in every organization

Create the necessary roles in all accounts:

aws cloudformation create-stack-instances --stack-set-name member-iam-access-role \ 
--profile root-org --accounts '["xxxxxxyyyyyy","yyyyzzzzzzz"]' \ 
--regions '["eu-west-1"]' --operation-preferences FailureToleranceCount=0,MaxConcurrentCount=1 \
--query 'OperationId'

where "xxxxxxyyyyyy" is the root account, "yyyyzzzzzzz" is one of the member accounts and the list is continued with all member accounts. Check the execution result (use the ID returned in the previous step as a parameter:

aws cloudformation describe-stack-set-operation \ 
--stack-set-name member-iam-access-role --profile root-org \ 
--operation-id xxx-yyyy-zzzzzz --query 'StackSetOperation.Status'

If you see "RUNNING", wait until it turns into "FAILED" or "SUCCEEDED". If failed, check the status with:

aws cloudformation list-stack-instances --profile root-org --stack-set-name member-iam-access-role

Now let's add a bit of bash-magic and add the SAML metadata to our final CloudFormation template. At this step the SAML provider (eg. Azure Enterprise Application) needs to be set up already. The metadata.xml is the SAML file downloaded in a previous step from the SAML Provider (Azure in this case).

sed -e "s/<MetadataDocument>/$(sed 's:/:\\/:g' ./metadata.xml)/" cfn_templates/saml-roles.yaml > cfn_templates/saml-roles-out.yaml
aws cloudformation create-stack-set --stack-set-name saml-providers-and-roles --profile root-org \
--capabilities CAPABILITY_NAMED_IAM \
--template-body file://./cfn_templates/saml-roles-out.yaml
aws cloudformation list-stack-sets --profile root-org --query "Summaries[?(@.StackSetName=='saml-providers-and-roles')].Status | [0]"

If Active we can install the IDP/SAML provider in all accounts.

aws cloudformation create-stack-instances --stack-set-name saml-providers-and-roles \
--profile root-org --accounts '["xxxxxxyyyyyy","yyyyzzzzzzz"]' \
--regions '["eu-west-1"]' --operation-preferences FailureToleranceCount=0,MaxConcurrentCount=1 \
--query 'OperationId'

Check the results (you should see SUCCEEDED):

aws cloudformation describe-stack-set-operation \
--stack-set-name saml-providers-and-roles --profile root-org \
--operation-id 17d21f5c-1a92-480f-9ac5-37ffeb30e60d --query 'StackSetOperation.Status'

Deploying the role synchronizing lambda function

Let's put the cherry on the cake.

The following command builds the lambda function and will run it on your local computer:

sam build && sam local invoke "RoleSyncFunction" -e schedule_evt.json --profile root-org --region eu-west-1

Now let's deploy it

Firstly, we need a S3 bucket where we can upload our Lambda functions packaged as ZIP before we deploy anything - If you don't have a S3 bucket to store code artifacts then this is a good time to create one:

aws s3 mb s3://role-sync-lmbda --profile root-org

Next, run the following command to package our Lambda function to S3:

sam package \
--output-template-file packaged.yaml \
--s3-bucket role-sync-lmbda --profile root-org

Next, the following command will create a Cloudformation Stack and deploy your SAM resources.

sam deploy \
--template-file packaged.yaml \
--stack-name role-sync-func \
--capabilities CAPABILITY_IAM --profile root-org

See Serverless Application Model (SAM) HOWTO Guide for more details in how to get started.

After deployment is complete you can run the following command to retrieve the API Gateway Endpoint URL:

Fetch, tail, and filter Lambda function logs

To simplify troubleshooting, SAM CLI has a command called sam logs. sam logs lets you fetch logs generated by your Lambda function from the command line. In addition to printing the logs on the terminal, this command has several nifty features to help you quickly find the bug.

NOTE: This command works for all AWS Lambda functions; not just the ones you deploy using SAM.

sam logs -n -RoleSyncFunction -stack-name role_sync --tail

You can find more information and examples about filtering Lambda function logs in the SAM CLI Documentation.

Cleanup

In order to delete our Serverless Application recently deployed you can use the following AWS CLI Command:

aws cloudformation delete-stack --stack-name role_sync

automate-saml-federation's People

Contributors

amazon-auto avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.