In order to better support our customers, we need to understand how they are actually using our products. This data should also help in the future if we decide to move towards a managed service offering. We're initially focusing on the product itself but later, we intend to expand this capability to cross product offerings such as Quarkus, OpenJDK and etc. 

The Jaeger Operator is installed as a prerequisite of Red Hat Service Mesh for OpenShift, and as such, we don’t have a clear view on how Jaeger is being used, and as such, how much value we are bringing to the product.

Given that we might not have the appropriate data in place, we need to better instrument the Jaeger Operator to provide information such as:

• How many Jaeger instances (CRs) are managed per operator
• Which strategies are being used (all in one, production, streaming)
• Which agent modes are being used (daemonset, sidecar)
• Which storages are being used (ES is the only supported one, but are people actually using others?)

We need to start getting answers to these questions by 2021-08-01, which means that we need to instrument the Jaeger Operator and release a set of patch/minor versions before that so that the Product Manager is able to get familiar with the data sources and required tooling.

Note that this goal requires effort from both engineering and product management.