Skip to content

list_managed_workgroups

list_managed_workgroups(
    *,
    config_overrides: Optional[RedshiftServerlessClientConfig] = None,
    source_arn: Optional[SourceArn] = None,
    next_token: Optional[PaginationToken] = None,
    max_results: Optional[int] = None,
) -> ListManagedWorkgroupsResponse

Returns information about a list of specified managed workgroups in your account.

Parameters:

  • source_arn (Optional[SourceArn], default: None ) –

    The Amazon Resource Name (ARN) for the managed workgroup in the Glue Data Catalog.

  • next_token (Optional[PaginationToken], default: None ) –

    If your initial ListManagedWorkgroups operation returns a nextToken, you can include the returned nextToken in following ListManagedWorkgroups operations, which returns results in the next page.

  • max_results (Optional[int], default: None ) –

    An optional parameter that specifies the maximum number of results to return. You can use nextToken to display the next page of results.

Raises:

Source code in docs_src/capo_redshift_serverless/_services/redshift_serverless.py
def list_managed_workgroups(
    self,
    *,
    config_overrides: Optional[RedshiftServerlessClientConfig] = None,
    source_arn: Optional[
        "capo_redshift_serverless.types.source_arn.SourceArn"
    ] = None,
    next_token: Optional[
        "capo_redshift_serverless.types.pagination_token.PaginationToken"
    ] = None,
    max_results: Optional[int] = None,
) -> "capo_redshift_serverless.types.list_managed_workgroups_response.ListManagedWorkgroupsResponse":
    """<p>Returns information about a list of specified managed workgroups in your account.</p>

    Args:
        source_arn: <p>The Amazon Resource Name (ARN) for the managed workgroup in the Glue Data Catalog.</p>
        next_token: <p>If your initial ListManagedWorkgroups operation returns a nextToken, you can include the returned nextToken in following ListManagedWorkgroups operations, which returns results in the next page.</p>
        max_results: <p>An optional parameter that specifies the maximum number of results to return. You can use nextToken to display the next page of results.</p>

    Raises:
        capo_redshift_serverless.errors.access_denied_exception.AccessDeniedException: <p>You do not have sufficient access to perform this action.</p>
        capo_redshift_serverless.errors.internal_server_exception.InternalServerException: <p>The request processing has failed because of an unknown error, exception or failure.</p>
        capo_redshift_serverless.errors.UnknownServiceError: The service returned an error code this client does not model.
    """

    def _handler(
        req: "OperationRequest[capo_redshift_serverless.types.list_managed_workgroups_request.ListManagedWorkgroupsRequest]",
    ) -> OperationResponse[
        "capo_redshift_serverless.types.list_managed_workgroups_response.ListManagedWorkgroupsResponse"
    ]:
        import capo_redshift_serverless._operations.redshift_serverless.list_managed_workgroups

        output, http_response = (
            capo_redshift_serverless._operations.redshift_serverless.list_managed_workgroups.list_managed_workgroups(
                req.options, req.input
            )
        )
        return OperationResponse(output=output, response=http_response)

    interceptors_, options_ = self.operation_options(
        config_overrides, OperationLogging("ListManagedWorkgroups")
    )
    input_: capo_redshift_serverless.types.list_managed_workgroups_request.ListManagedWorkgroupsRequest = {}
    if source_arn is not None:
        input_["source_arn"] = source_arn
    if next_token is not None:
        input_["next_token"] = next_token
    if max_results is not None:
        input_["max_results"] = max_results

    response = execute_pipeline(
        OperationRequest(input=input_, options=options_),
        handler=_handler,
        interceptors=list(interceptors_),
    )
    response.response.close()
    return response.output