# cdk-sqlserver-seeder [](https://github.com/eladb/awesome-cdk)



[](https://badge.fury.io/js/cdk-sqlserver-seeder)
[](https://badge.fury.io/py/cdk-sqlserver-seeder)
[](https://badge.fury.io/nu/Talnakh.SqlServerSeeder)
[](https://repo1.maven.org/maven2/xyz/talnakh/SqlServerSeeder/)
A simple CDK seeder for SQL Server RDS databases.
When you create an RDS SQL Server instance using CloudFormation template, there is no way to provide initial
schema definition as part of CloudFormation stack deployment. Custom schema deployment scripts can be executed
only after the database deployment is complete.
*cdk-sqlserver-seeder* library is a [AWS CDK](https://aws.amazon.com/cdk/) construct that provides a way to automate
this process and eliminate manual steps involved in the process of preparing new RDS SQL Server environment by
executing custom SQL scripts on RDS SQL Server instance creation/deletion.
The construct relies on [Invoke-SqlCmd](https://docs.microsoft.com/en-us/powershell/module/sqlserver/invoke-sqlcmd) cmdlet
to run the scripts and provides a way to handle transient errors during stack provisioning.
## Usage
```python
# Example automatically generated without compilation. See https://github.com/aws/jsii/issues/826
import aws_cdk.core as cdk
import aws_cdk.aws_ec2 as ec2
import aws_cdk.aws_rds as rds
from cdk_sqlserver_seeder import SqlServerSeeder
class DatabaseStack(cdk.Stack):
def __init__(self, scope, id, *, description=None, env=None, stackName=None, tags=None, synthesizer=None, terminationProtection=None):
super().__init__(scope, id, description=description, env=env, stackName=stackName, tags=tags, synthesizer=synthesizer, terminationProtection=terminationProtection)
sql_server = rds.DatabaseInstance(self, "Instance",
engine=rds.DatabaseInstanceEngine.SQL_SERVER_WEB
)
seeder = SqlServerSeeder(self, "SqlSeeder",
database=sql_server,
port=1433,
vpc=vpc,
create_script_path="./SQL/v1.0.0.sql", # script to be executed on resource creation
delete_script_path="./SQL/cleanup.sql"
)
```
## Configuration properties
SqlServerSeeder construct accepts the following configuration properties:
| Parameter | Required | Default | Description |
|---|---|---|---|
| `vpc` | yes | | VPC for Lambda function deployment |
| `database` | yes | | RDS SQL Server database instance |
| `createScriptPath` | yes | | SQL scripts to run on resource creation |
| `deleteScriptPath` | no | | SQL script to run on resource deletion |
| `port` | no | 1433 | RSD SQL Server database port |
| `memorySize` | no | 512 | Lambda function memory size |
| `ignoreSqlErrors` | no | false | Whether to ignore SQL error or not |
## Architecture

`cdk-sqlserver-seeder` deploys a custom resource backed by PowerShell lambda to connect to SQL Server instance. Lambda function is deployed in private subnets of your VPC where RDS instance resides.
Lambda function retrieves database credentials from [AWS Secrets Manager](https://aws.amazon.com/secrets-manager/) and uses them to construct connection string to the database.
SQL scripts are uploaded into S3 bucket during CDK application deployment.
Lambda function downloads these scripts during execution.
## Security considerations
Lambda function has the following permissions:
* Managed policies
* `AWSLambdaBasicExecutionRole` for CloudWatch logs
* `AWSLambdaVPCAccessExecutionRole` for VPC access
* Inline policy
* `secretsmanager:GetSecretValue` for RDS credentials secret
* `s3:GetObject*`, `s3:GetBucket*`, `s3:List*` for S3 bucket with SQL scripts
## Acknowledgements
The whole project inspired by [aws-cdk-dynamodb-seeder](https://github.com/elegantdevelopment/aws-cdk-dynamodb-seeder).
I though it would be very helpful to have a similar way to seed initial schema to more traditional SQL Server databases.