This is a simple bootstrapper for running the Elasticsearch Curator as an AWS Lambda-function. Lambda functions can be triggered using CloudWatch Events making it a suitable environment for running Curator tasks.

Note: Currently Lambda functions have a maximum execution time of 300 seconds so long-running Curator tasks are not suitable for running as Lambda functions.

• ES_HOST: the base url of the Elasticsearh cluster (default: http://localhost).
• ES_PORT: the Elasticseach cluster port (default: 9200).
• DRY_RUN: set Curator --dry-run flag (default: True)
• FUNCTION_NAME: the lambda function name (default: lambda-curator).
• IAM_ROLE: the lambda function execution role.

The AWS CLI needs to be installed in order to push/update the lambda-curator function. As a minimum the following variables need to be set:

• AWS_ACCESS_KEY_ID: the aws access key id.
• AWS_SECRET_ACCESS_KEY: the aws secret access key.
• AWS_REGION: the aws region.

Add an actions.yml file with the actions to be triggered by the lambda function.

make package

This will create a FUNCTION_NAME.zip including the bootstrap.py (and all Curator dependencies) and actions.yml.

make awspush

This will create and publish the lambda-curator function.

make awsupdate

This will update any existing lambda-curator function.

The lambda-curator function will sign all requests with the AWS credentials of the execution role (IAM_ROLE). Giving this role access to your AWS Elasticsearch cluster using IAM will authenticate all lambda-curator actions automagically.