Use CreateDeployment with an AWS SDK or CLI

Example 1: To create a deployment

The following create-deployment example deploys the AWS IoT Greengrass Command Line Interface to a core device.

aws greengrassv2 create-deployment \
    --cli-input-json file://cli-deployment.json

Contents of cli-deployment.json:

{
    "targetArn": "arn:aws:iot:us-west-2:123456789012:thing/MyGreengrassCore",
    "deploymentName": "Deployment for MyGreengrassCore",
    "components": {
        "aws.greengrass.Cli": {
            "componentVersion": "2.0.3"
        }
    },
    "deploymentPolicies": {
        "failureHandlingPolicy": "DO_NOTHING",
        "componentUpdatePolicy": {
            "timeoutInSeconds": 60,
            "action": "NOTIFY_COMPONENTS"
        },
        "configurationValidationPolicy": {
            "timeoutInSeconds": 60
        }
    },
    "iotJobConfiguration": {}
}

Output:

{
    "deploymentId": "a1b2c3d4-5678-90ab-cdef-EXAMPLE11111"
}

For more information, see Create deployments in the AWS IoT Greengrass V2 Developer Guide.

Example 2: To create a deployment that updates component configurations

The following create-deployment example deploys the AWS IoT Greengrass nucleus component to a group of core devices. This deployment applies the following configuration updates for the nucleus component:

Reset the target devices' proxy settings to their default no proxy settings.Reset the target devices' MQTT settings to their defaults.Sets the JVM options for the nucleus' JVM.Sets the logging level for the nucleus.

aws greengrassv2 create-deployment \
    --cli-input-json file://nucleus-deployment.json

Contents of nucleus-deployment.json:

{
    "targetArn": "arn:aws:iot:us-west-2:123456789012:thinggroup/MyGreengrassCoreGroup",
    "deploymentName": "Deployment for MyGreengrassCoreGroup",
    "components": {
        "aws.greengrass.Nucleus": {
            "componentVersion": "2.0.3",
            "configurationUpdate": {
                "reset": [
                    "/networkProxy",
                    "/mqtt"
                ],
                "merge": "{\"jvmOptions\":\"-Xmx64m\",\"logging\":{\"level\":\"WARN\"}}"
            }
        }
    },
    "deploymentPolicies": {
        "failureHandlingPolicy": "ROLLBACK",
        "componentUpdatePolicy": {
            "timeoutInSeconds": 60,
            "action": "NOTIFY_COMPONENTS"
        },
        "configurationValidationPolicy": {
            "timeoutInSeconds": 60
        }
    },
    "iotJobConfiguration": {}
}

Output:

{
    "deploymentId": "a1b2c3d4-5678-90ab-cdef-EXAMPLE11111",
    "iotJobId": "a1b2c3d4-5678-90ab-cdef-EXAMPLE22222",
    "iotJobArn": "arn:aws:iot:us-west-2:123456789012:job/a1b2c3d4-5678-90ab-cdef-EXAMPLE22222"
}

For more information, see Create deployments and Update component configurations in the AWS IoT Greengrass V2 Developer Guide.

    def create_deployment(
        self,
        target_arn: str,
        deployment_name: str,
        components: dict[str, Any],
        deployment_policies: Optional[dict[str, Any]] = None,
    ) -> dict[str, Any]:
        """
        Creates a deployment targeting a thing group or individual device.

        :param target_arn: The ARN of the target thing or thing group.
        :param deployment_name: A human-readable name for the deployment.
        :param components: A map of component names to deployment configurations.
        :param deployment_policies: Optional deployment policies (failure handling, etc.).
        :return: A dictionary containing the deployment ID and IoT job details.
        """
        try:
            params = dict()
            params["targetArn"] = target_arn
            params["deploymentName"] = deployment_name
            params["components"] = components
            if deployment_policies is not None:
                params["deploymentPolicies"] = deployment_policies
            response = self.client.create_deployment(**params)
            logger.info(
                "Created deployment %s targeting %s.",
                response.get("deploymentId"),
                target_arn,
            )
            return response
        except ClientError as err:
            if err.response["Error"]["Code"] == "ValidationException":
                logger.error(
                    "Invalid deployment parameters. Check that the targetArn is a valid "
                    "thing or thing group ARN and component versions exist. %s",
                    err.response["Error"]["Message"],
                )
            raise