Currently we don't make a clear recommendation about which properties in the YAML should really be configured. As https://issues.jboss.org/browse/ENTMQST-971 demonstrates this means users can end up having problems because they were unaware of configuration which we consider recommended.

It's not always possible to add this to the the examples (for instance where the right values might depend on workload, or the platform).

So we should do a better job of signposting this in the documentation. While we might use `IMPORTANT` in the docs that means the user has to read a lot of the documentation to know all the things we recommend, so it is less than ideal.

One way to do this might be starting section "3.1. Kafka cluster configuration" with a skeleton Kafka YAML which included all the properties we recommend ought to be configured. Callouts could be used with links to the more in-depth documentation. That would provide a good, and very visual, overview and make it easier for users to check they had all the recommended configs set by comparing with their resources.

This should include:

• ZK and Kafka Storage
• ZK and Kafka Replicas
• ZK and Kafka memory and cpu requests (including sidecars, I suppose)
• Rack awareness

Anything else scholzj?