What do you need to definitively answer the question of is your application working, and furthermore is it working well? Observability: Consisting of at least the logs, metrics, and traces of your services and potentially additional data such as availability and response time from multiple geographies, the load balancer and firewall protecting and serving your application, packet-level tracing between services, and security information you can get a complete picture of what is happening inside, between, around, and to your application. You can collect all of this information however your application runs, on a server, vm, cloud, container, or serverless. But to make sense of all that data you've got to get it into a place where you can search, correlate, and pivot between sources to rapidly identify the root cause of issues in real-time, search for leading indicators before application impacts to achieve a proactive rather than reactive response to issues, and have the historical view you need to make decisions about trending and capacity planning.