[Feature Request] Allow Prometheus ServiceMonitor to target vault-internal service in HA mode

[krarey]

Problem:
The ServiceMonitor resource that is created by this chart will only target the vault-active service if Vault is deployed in HA mode. Nodes in the Standby and Performance Standby state emit metrics that are useful when diagnosing cluster performance and request errors, which are not collected in this configuration.

Solution:
Add a flag or other logic which allows the user to target the vault-internal service, even when HA mode is enabled. This will let Prometheus scrape all Vault pods in the service's endpoint list to collect standby metrics. This should be accompanied by a note that unauthenticated metrics access should be enabled in this configuration, since non-performance standbys cannot otherwise serve the request.