Add event addition metrics for latency and counter

[stevend-uber]

Currently in the event-based cache hydration feature, we have metrics around latency/counts for fetch and prunes of events from the events table. I propose we add a counter and latency metrics around Event-Addition.

[Victorblsilveira]

Most of the metrics creation is done at metricsWrapper. Another aspect of this is that event creation is done in createRegistrationEntryEvent, an internal function. Even if we add the function to the interface and implement the metrics start at metricsWrapper, most internal calls wouldn't fire the latency/count metric.

One solution would be to internalize the metrics fire to the sqlstore.go. Although looking at the code this breaks the patterns. Is that an acceptable solution for the spire community? I want to help here and appreciate any guidance.

[rturner3]

@Victorblsilveira Your suggested approach seems reasonable to me.

The only alternative I could think of would be to refactor the datastore interface to make the creation of events an explicit call from the entry cache manager. But that's not really ideal because it introduces the possibility of business logic bugs where events are mistakenly not written to the database table for write/delete operations to entries/attested nodes.

[azdagron]

Talking in the contributor sync, we determined that with #4720 we should have enough metrics to start with for the events based cache.