This CDK script deploys the following the components:
- VPC for MSK Serverless and Kinesis Data Analytics application.
- MSK Serverless.
- Kinesis Data Analytics Python Table API application.
- IAM permissions for the role associated with the Kinesis Data Analytics application.
This CDK script expects you to supply the following existing resources:
- S3 bucket where the application jar will be uploaded (
appBucket
below). - S3 bucket that will function as the sink (
appSinkBucket
below). - Glue database (
glueDatabaseName
below).
Open up cdk.json
and fill in appropriate values for each of these CDK context values:
Context value name | Purpose | Notes |
---|---|---|
kdaAppName |
The name of the Kinesis Data Analytics application | KDA app will be created |
appBucket |
The S3 bucket where the application payload will be stored | Must be pre-existing |
appSinkBucket |
The bucket to which the MSK to S3 Flink app will write output files (in Parquet) | Must be pre-existing |
runtimeEnvironment |
The Kinesis Data Analytics runtime environment | For instance, FLINK-1_15 |
deployDataGen |
true if you want Zeppelin-based interactive KDA for data generation to be deployed; false otherwise |
N/A |
glueDatabaseName |
The AWS Glue database that will be used by KDA Studio datagen app | Must be pre-existing |
kdaLogGroup |
The name for the CloudWatch Log Group that will be linked to the KDA Flink app | Log group will be created |
kdaLogStream |
The name for the CloudWatch Log Stream that will be linked to the KDA Flink app | Log stream will be created |
sourceMskClusterName |
The name for the source MSK Serverless cluster | MSK Serverless cluster will be created |
For more information on CDK Runtime Context, please see Runtime Context.
cdk deploy
This will launch a CloudFormation Stack containing all the resources required for the blueprint.
Instead of deploying directly, you could also generate an intermediate CFN script using the command below.
cdk synth
To avoid ongoing charges, please make sure that you delete the blueprint and associated AWS resources using the following command.
cdk destroy