So by attempting to use Datadog to help track and thereby offset some of the costs of what would be considered minor data ended up driving up costs due to the sheer number of these requests? Is this a problem that is more centered around serverless?