The default setting for the Dataset name options is generic (see the screenshot below).

However, with any dataset names other than aws.cloudwatch_logs, the custom pipeline will not work. Fleet always creates the index template with an index pattern logs-aws.cloudfront_logs-* that only matches the default integration dataset (aws.cloudwatch_logs) and will not match the custom dataset name provided.

On integration packages, Fleet always uses the default integration dataset. Fleet uses the custom dataset name only with input packages.

To avoid having a non-working custom pipeline OOTB, we can:

• Change the predefined dataset name from generic to aws.cloudwatch_logs to make the custom pipeline work OOTB.
• Add a disclaimer to the dataset name config option to suggest they must create an index template manually.