Alerting monitor clients when log STH hasn't been updated for a while

[efebarlas]

This feature can be accomplished either in the infrastructure (a cron job or something similar runs the poll job; if n periods of the poll job never alerted the cron job that the poll found an updated STH the user is alerted via a piece of go code), or in the go code itself (for loop polls the log periodically and counts by itself).

I think taking care of this in the infrastructure makes more sense because the jobs themselves should be quick after they are run by the monitor, so periodic checks inside Go might not be what we're looking for.

What is your opinion on this? Also, I can work with someone from the infrastructure team on polling, the Go code required is almost done.

[haydentherapper]

Now with stable checkpoints returned from Rekor, we will want this. We'll call this "checkpoint freshness", and a Verifier should alert if it remains stale for some time.