Came up with an initial design for retry mechanism based off of all the prior conversation. First, some assumptions: 1) Rules that currently would benefit from a retry mechanism:

test

and

url_to_fetch

2) In the future, any rule that may fail due to external environment should be retryable, such as if we add an

HttpRequest

rule 3) Given 1 and 2, we want a generic mechanism to say that a rule should be retryable. We should not special case

test

4) The retry mechanism must be hooked up to options, such as

--test-pytest-max-retries