Created Job names can collide when being in an unfinished state

[pcktdmp]

Currently Jobs created by cnfuzz do not properly cleanup old or previously created Jobs.
This becomes a problem when Jobs are being re-created, for example upon re-installation of cnfuzz.

[Luukvdm]

cnfuzz might start fuzzing an API, and if the API updates during the fuzzing process you might also run into this bug.
This raises the question of whether the existing job should be stopped because it might no longer be relevant.

[pcktdmp]

It should stop and clear it's cache entry. Then the case is covered that a new api version is going to be fuzzed and if no new version is being deployed the fuzzer will (keep) retrying until completion when the pod is being rescheduled (or not).