Retina Windows Agent should (non-silently) fail if map callback registration fails

[matthewige]

Describe the bug
Currently, only 1 consumer of the EVENTS_MAP can exist at a time on the system. If another process has already register, the callback in the retina windows agent currently silently fails. This can lead to a false-positive situation, where the pod / agent appears healthy, but it is not actually collecting the required metrics.

To Reproduce
Have a consume of the EVENTS_MAP prior to starting the retina windows pod

Expected behavior
Error message (possibly failed pod startup) occurs, to be noisy about the failure.

Screenshots
n/a
Platform (please complete the following information):

• OS: WindowsAnnual
• Kubernetes Version: [e.g. 1.22]
• Host: [e.g. AKS, KIND, self-host, etc]
• Retina Version:

Additional context
n/a