Apologies for the delayed response on this.
(1) I assume there is no problem with creating a kubernetes service and kubernetes end points as described here.
- Yes the approach followed is correct.
(2) Can someone explain how to create a kubernetes reliable singleton (as described using mysql in kubernetes-up-and-running-2nd-edition chapter 15) using azure serverless SQL instead of mysql? Basically the authors show how to create a persistent volume & claim connected to a replicaset of 1 for mysql.
- If its singleton, Then its not reliable. Running 1 replica is not recommended
(3) Can someone explain how to create a kubernetes stateful set using serverless azure SQL?
- As SQL is a managed service, this is not possible. Instead you can create a sql separately and then add that as a headless service in Kubernetes.
Please find this document that has instructions on deploying a SQL server container in Kubernetes with AKS.