Hi @Charlotte Brady :) You can easily create your own fetch provider (by inheriting from the existing one, or from scratch) and add more logging and observability there. That being said, I would instead recommend a different direction. As long as the data source service is alive and well, you shouldn't care if there are intermittent failures, that's the beauty of the retry mechanisms in OPAL. And you can instead perform independent health checks on the data source service. This would produce a much more decoupled, k8s style, and clearer (with less nested logs) monitoring events.