StartBulkDeploymentRequest
¶
Source code in docs_src/capo_greengrass/types/start_bulk_deployment_request.py
amzn_client_token
instance-attribute
¶
amzn_client_token: NotRequired[__string]
A client token used to correlate requests and responses.
execution_role_arn
instance-attribute
¶
execution_role_arn: NotRequired[__string]
The ARN of the execution role to associate with the bulk deployment operation. This IAM role must allow the ''greengrass:CreateDeployment'' action for all group versions that are listed in the input file. This IAM role must have access to the S3 bucket containing the input file.
input_file_uri
instance-attribute
¶
input_file_uri: NotRequired[__string]
The URI of the input file contained in the S3 bucket. The execution role must have ''getObject'' permissions on this bucket to access the input file. The input file is a JSON-serialized, line delimited file with UTF-8 encoding that provides a list of group and version IDs and the deployment type. This file must be less than 100 MB. Currently, AWS IoT Greengrass supports only ''NewDeployment'' deployment types.