Creating a CDK resource in TypeScript if it doesn't already exist

Question

Creating a CDK resource in TypeScript if it doesn't already exist

Recently, I set up a DynamoDB table in my CDK project to be used by lambdas within the same project. Everything was working fine until we needed to delete the stack with the retain resource option set to true for the table.

However, when attempting a fresh deployment, we encountered an error stating that the table already exists, leading to the entire stack rolling back. I am now looking for code that will create the table only if it does not already exist.

Here is a snippet of the basic table creation code. Despite searching extensively, I have been unable to find any documentation addressing this issue or detailing the specific exception that needs to be caught. All we see are logs on the AWS CloudFormation console.

 const dynamoTable = new Table(this, "my-table", {
      tableName: StackConfiguration.tableName,
      partitionKey: { name: "id", type: AttributeType.STRING },
    });

typescript amazon-web-services amazon-dynamodb aws-cdk

Answer 1

Answer №1

Regrettably, accomplishing that task in CDK is not possible due to the fact that CDK generates CloudFormation template during compile time rather than runtime. There are a few alternative approaches available:

Utilize CloudFormation Resource Import to bring an existing table into your stack
Implement a Custom resource lambda function to execute an AWS API call for checking the existence of the table. Then use the Custom resource output within Fn.conditionEquals in the CDK code to conditionally create the table

If this task is a one-time occurrence, it is suggested to go with the first option. On the other hand, if this task is expected to be recurring, option 2 would be more suitable.

Answer 2

Regrettably, accomplishing that task in CDK is not possible due to the fact that CDK generates CloudFormation template during compile time rather than runtime. There are a few alternative approaches available:

Utilize CloudFormation Resource Import to bring an existing table into your stack
Implement a Custom resource lambda function to execute an AWS API call for checking the existence of the table. Then use the Custom resource output within Fn.conditionEquals in the CDK code to conditionally create the table

If this task is a one-time occurrence, it is suggested to go with the first option. On the other hand, if this task is expected to be recurring, option 2 would be more suitable.

Answer 3

Answer №2

Starting in September 2023, the CloudFormation Development Kit (CDK) now offers the ability to import existing resources. I successfully followed the steps outlined to import my existing DynamoDB table using the CDK. However, upon running the cdk import command, an error was thrown stating that there were changes detected in my Lambda functions despite the cdk diff showing no modifications. To resolve this issue, I had to include the --force option with the import command.

Please note that this method did not align with the original intention of @AnonymousAlias, who wanted to either create the resource if it didn't exist or skip it if it did. As suggested by @Tofig Hasanov, a workaround would involve checking for the existence of the resource before creating it.

Referencing https://docs.aws.amazon.com/cdk/v2/guide/cli.html#cli-import:

Importing existing resources into a stack

The cdk import command allows you to incorporate resources into the management of an AWS CDK stack through CloudFormation. This is useful for scenarios such as transitioning to AWS CDK, migrating resources between stacks, or altering their logical identifier. The cdk import operation leverages CloudFormation resource imports. Here's a list of supported resources eligible for import.

To import an existing resource into an AWS CDK stack, follow these steps:

Ensure the resource is not under the governance of another CloudFormation stack. If so, adjust the removal policy to RemovalPolicy.RETAIN within the current stack, conduct a deployment, then exclude the resource and deploy again to release it from CloudFormation management without deletion.

Perform a cdk diff operation to confirm no impending changes in the target AWS CDK stack. Only additions of new resources are permitted during an "import" process.

Add constructs representing the resources intended for import into your stack. For instance, to import an Amazon S3 bucket, add a construct like new s3.Bucket(this, 'ImportedS3Bucket', {});. Avoid making adjustments to other resources.

It's crucial to accurately model the current state of the resource into the definition. Include all pertinent details such as AWS KMS keys and lifecycle policies for the bucket. Neglecting this could lead to unexpected outcomes during subsequent updates.

You have the flexibility to decide whether or not to include the physical bucket name. Omitting resource names from your AWS CDK definitions simplifies the process for deploying resources repeatedly.

Execute cdk import STACKNAME.

If the resource names are absent in your model, the CLI will prompt you to provide the actual names of the resources being imported. Once identified, the import process commences.

Upon successful completion reported by cdk import, the resource becomes managed by AWS CDK and CloudFormation. Subsequent alterations made to the resource properties within your AWS CDK app or the construct configuration will take effect in the next deployment.

To verify that the resource definition in your AWS CDK app corresponds with the current state of the resource, initiate a CloudFormation drift detection operation.

Kindly note that this feature currently does not support importing resources into nested stacks.

Answer 4

Starting in September 2023, the CloudFormation Development Kit (CDK) now offers the ability to import existing resources. I successfully followed the steps outlined to import my existing DynamoDB table using the CDK. However, upon running the cdk import command, an error was thrown stating that there were changes detected in my Lambda functions despite the cdk diff showing no modifications. To resolve this issue, I had to include the --force option with the import command.

Please note that this method did not align with the original intention of @AnonymousAlias, who wanted to either create the resource if it didn't exist or skip it if it did. As suggested by @Tofig Hasanov, a workaround would involve checking for the existence of the resource before creating it.

Referencing https://docs.aws.amazon.com/cdk/v2/guide/cli.html#cli-import:

Importing existing resources into a stack

The cdk import command allows you to incorporate resources into the management of an AWS CDK stack through CloudFormation. This is useful for scenarios such as transitioning to AWS CDK, migrating resources between stacks, or altering their logical identifier. The cdk import operation leverages CloudFormation resource imports. Here's a list of supported resources eligible for import.

To import an existing resource into an AWS CDK stack, follow these steps:

Ensure the resource is not under the governance of another CloudFormation stack. If so, adjust the removal policy to RemovalPolicy.RETAIN within the current stack, conduct a deployment, then exclude the resource and deploy again to release it from CloudFormation management without deletion.

Perform a cdk diff operation to confirm no impending changes in the target AWS CDK stack. Only additions of new resources are permitted during an "import" process.

Add constructs representing the resources intended for import into your stack. For instance, to import an Amazon S3 bucket, add a construct like new s3.Bucket(this, 'ImportedS3Bucket', {});. Avoid making adjustments to other resources.

It's crucial to accurately model the current state of the resource into the definition. Include all pertinent details such as AWS KMS keys and lifecycle policies for the bucket. Neglecting this could lead to unexpected outcomes during subsequent updates.

You have the flexibility to decide whether or not to include the physical bucket name. Omitting resource names from your AWS CDK definitions simplifies the process for deploying resources repeatedly.

Execute cdk import STACKNAME.

If the resource names are absent in your model, the CLI will prompt you to provide the actual names of the resources being imported. Once identified, the import process commences.

Upon successful completion reported by cdk import, the resource becomes managed by AWS CDK and CloudFormation. Subsequent alterations made to the resource properties within your AWS CDK app or the construct configuration will take effect in the next deployment.

To verify that the resource definition in your AWS CDK app corresponds with the current state of the resource, initiate a CloudFormation drift detection operation.

Kindly note that this feature currently does not support importing resources into nested stacks.

Answer 5

Answer №3

Although not the most optimal solution, I have come up with a workaround that might be helpful to someone. By incorporating table creation within a try-catch block in our code, we can handle exceptions more effectively. Instead of catching a general exception, it would be useful to identify the specific one for better error handling. This approach ensures the successful deployment of the stack.

 try {
     const dynamoTable = new Table(this, "my-table", {
      tableName: StackConfiguration.tableName,
      partitionKey: { name: "id", type: AttributeType.STRING },
    });
    
      return dynamoReplayTable;
    } catch (e) {
      return;
    }

To utilize the table effectively in your code, consider referencing the ARN instead of the table variable name. Alternatively, you could explore importing from the name within the catch block. However, the most efficient solution is to maintain tables in a separate stack.

Answer 6

Although not the most optimal solution, I have come up with a workaround that might be helpful to someone. By incorporating table creation within a try-catch block in our code, we can handle exceptions more effectively. Instead of catching a general exception, it would be useful to identify the specific one for better error handling. This approach ensures the successful deployment of the stack.

 try {
     const dynamoTable = new Table(this, "my-table", {
      tableName: StackConfiguration.tableName,
      partitionKey: { name: "id", type: AttributeType.STRING },
    });
    
      return dynamoReplayTable;
    } catch (e) {
      return;
    }

To utilize the table effectively in your code, consider referencing the ARN instead of the table variable name. Alternatively, you could explore importing from the name within the catch block. However, the most efficient solution is to maintain tables in a separate stack.

Creating a CDK resource in TypeScript if it doesn't already exist

Answer №1

Answer №2

Importing existing resources into a stack

Answer №3

Similar questions

Ways to display Leaflet pins within Angular?

Attempting to utilize Array Methods with an Array Union Type

Declaration of types for invoking the lodash `mapKeys` function

"Optimizing Performance: Discovering Effective Data Caching

The display of the selected input is not appearing when the map function is utilized

When using Angular, receiving a "Bad Request" error message with a status code of 400 after making a POST

Best practices for implementing dual ngFor directives within a single tr element?

Building a gateway for integrating a scikit-learn model within AWS Sagemaker

Steering clear of the generic Function type in React with TypeScript

Turning an array of strings into a multidimensional array

Webpack 5 is failing to bundle re-exports correctly

Encountering difficulties transferring data from a parent to a child component within Angular 9

Utilize Immer and ES6 to effortlessly update a nested object within the Redux store by manipulating normalized data structure

Experiencing issues with effectively using a component

Can you provide some insight into why the init() method is throwing an error?

Modifying the value of a property in an object array created using the map method is ineffective

Where can I find the endpoint URL for the Node backend deployed to the AWS S3 bucket?

Issue "Module not found" arises while trying to import an external JSON file in TypeScript

Error 404 Encountered During Azure Functions GET Request

The issue arises when attempting to call a method from the same service within jsPDF in an Angular environment