fix: consider zero when retry_count not present

[IgorFroehner]

Fix Explanation

We're having some errors in the first failure of jobs with silent active because the job_payload won't have the retry_count key if there were no retries. This is happening because Sidekiq sets the retry_count key on the job_payload to zero only after the first failure of the job: see source. As a side note, observe that the first try doesn't count as a retry in the retry_count, so, the way it's implemented today, the warn_after means warn after # of retries and not # of tries (first try + retry_count) in general.

fix: #6