Refactor the config map routing loop in the Scaler

[arschles]

Currently, when the config map routing loop experiences an update, we do a full update from the k8s API using the scaler's queue pinger.

It may be more efficient to just update the pinger's in-memory copy the host list to reflect the change in the routing table (the pinger would still be pinging all interceptors in a separate goroutine), rather than making a new request to all interceptors every time a change to the set of HTTPScaledObjects is made.

This is a follow-up from #326

Use-Case

A reduction (however small) to the number of network requests made from scaler to interceptors.

Specification

When an event occurs on the HTTPScaledObject set, we should avoid issuing "fan-out" requests (i.e. scaler issues a metrics request to each interceptor and aggregates results) as a result.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in 7 days if no further activity occurs. Thank you for your contributions.

[arschles]

This has been completed in #326 by @tpiperatgod. closing