[Feature Request] Option to restrict read service to replica pods only (exclude Primary) for 100% read HA #136
Description
Currently, when deploying Valkey with replicas, the read service (e.g., my-release-valkey-read) seems to select all pods in the cluster, including the primary pod.
In the scenario where a client opens a read connection and lands on the primary pod, if the primary pod crashes, also read operations will fail although we have other two replicas available.
It would be nice to have a configuration option to configure the read service to strictly select only replica pods, effectively excluding the primary pod, in order to achieve at least a 100% HA for read operations in case of master crash.