Instance not active log is confusing

[huizhilu]

Description

In Helix REST, when we call stoppable api, helix own custom health check is performed. And there is a warn log:
[WARN ] 2020-10-17 11:45:19,355 org.apache.helix.util.InstanceValidationUtil - The instance <instance> is not active
And then custom instance health is also checked by calling the external custom REST.

So this is confusing: if the instance is checked to be inactive, why do we still call the custom instance health check?

Analysis

The check is actually in hasErrorPartitions(). If the instance doesn't have error partitions, the warn message is logged. This is inaccurate. We should remove this warn log.

[huizhilu]

Fixed by #1490