Skip to content

Commit

Permalink
cleaning up control flow doc
Browse files Browse the repository at this point in the history
  • Loading branch information
Sakthi Vetrivel committed Jun 25, 2020
1 parent 604d820 commit c00ff2e
Show file tree
Hide file tree
Showing 2 changed files with 14 additions and 20 deletions.
34 changes: 14 additions & 20 deletions docs/design/controlflow.md
Original file line number Diff line number Diff line change
@@ -1,47 +1,41 @@
## Deploying the operator - How it works
## Deploying the operator - how it works

![](/docs/images/DeployFlow.png)

1. The user provisions the Managed Identity (or service principal) in their environment for the Azure Service Operator to use.
2. The user then deploy cert-manager for the Azure Service Operator to use when deployed.

2. The user then uses the Helm chart to deploy the operator.
3. The user then uses the Helm chart to deploy the operator. This includes the controller pod and the manager pod, created from images stored in the public Microsoft Container Registry. The Helm chart also deploys dependencies like [aad-pod-identity](https://github.com/Azure/aad-pod-identity) in addition to the Azure Service Operator. This deployment also configures the needed Kubernetes RBAC rules.

3. The Helm chart uses the images from the public Container Registry.

4. It deploys dependencies like [aad-pod-identity](https://github.com/Azure/aad-pod-identity) in addition to the Azure Service Operator.
## Resource provisioning - how it works

5. This deployment takes care of deploying the operator while also configuring the needed Kubernetes RBAC rules.
![](/docs/images/ASO%20flow.png)

6. Due to ordering of deployment, currently, cert-manager needs to be deployed manually before using the Helm chart to deploy the Operator.
1. The user deploys an application that includes the custom resource manifest for installing a Azure service that the app depends on.

## Resource provisioning - How it works
2. The application is deployed using its manifest. However, the deployment does not yet succeed as it waits on the Azure service to be successfully created. The application references a secret that provides the information required to consume the Azure service, and the secret does not exist yet.

![](/docs/images/ProvisionFlow.png)
3. The Azure Service Operator continously watches the custom resource definitions (CRDs) corresponding to the Azure services and recognizes the request for a custom resource.

1. The developer deploys an application that includes the manifest for installing a Azure service that the app depends on.
4. The Azure Service Operator then updates the Kubernetes instance for the requested resource with the correct status and events.

2. The application is deployed using its manifest, however the deployment does not succeed yet as it is waiting for the Azure Service that it depends on. The application references a secret that provides the information required to consume the Azure service, and the secret does not exist yet.
5. The Azure Service Operator requests an authorizer from Azure Active Directory for the Azure resource management endpoint, as the identity it is running as and receives an authorizer token.

3. The Azure Service Operator that watches the custom resource definitions (CRDs) corresponding to the Azure services recognizes the request for a custom resource.

4. The Azure Service Operator updates the Kubernetes instance for the requested resource with the correct Status and events.

5. The Azure Service Operator requests an authorizer from Azure Active Directory for the Azure resource management endpoint, as the identity it is running as.

6. The Azure Service Operator then sends the provisioning request to Azure API, along with the authorizer in the request.
6. The Azure Service Operator then sends the provisioning request to Azure API, along with the authorizer token in the request.

7. Azure API provisions/deprovisions the resource and returns the Resource object to the Service Operator.

8. The Azure Service Operator retrieves the information required to access/consume the Azure resource from the Resource objecrt and stores it in a Kubernetes secret or as a secret in a specified Azure KeyVault.
8. The Azure Service Operator retrieves the information required to access/consume the Azure resource from the Resource object and stores it in a Kubernetes secret or as a secret in a pre-specified Azure KeyVault.

9. The app is deployed successfully now that the Azure service it depends on is provisioned, and the secret it references exists.

## Security considerations

1. The Azure Service Operator requests the authorizer from AAD for every provision/deprovision request. There is no caching of security tokens.

2. Running the Azure Service Operator under a Managed Identity is recommended for security reasons. There is support to use Service Principals if needed, but not recommended.
2. Running the Azure Service Operator under a Managed Identity is recommended for security reasons. Azure Service Operator can also be run using service principals, but this is not recommended.

3. It is recommended to use Azure KeyVault to store connection information, keys that are an output of the provisioning process. There is support to store as Kubernetes secrets but not recommended.
3. It is recommended to use Azure KeyVault to store connection information, AKA keys that are an output of the provisioning process. These connection strings can be stored as Kubernetes secrets, but this is not recommended.

4. There is no implicit deletion of resources. Resources will be deprovisioned/deleted only on an explicit delete operation.
Binary file added docs/images/ASO flow.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.

0 comments on commit c00ff2e

Please sign in to comment.