Add logic to capture multiple occurrences of exceptions within a timeframe
Issue #8
new
For a particular kind of exception, we should be able to set a watch that notifies someone (or flags that exception for review) if it occurs more than x times in y minutes.
Example: Some kinds of messages, like maybe a db timeout, are not necessarily a problem if they happen once a month, but if it happens twice in one minute, it may indicate a problem.