What is the nature and description of the request?

There was a feature added to receptor to allow it to retry in instance of errors communicating with the Kube API. This allowed it to be slightly more resilient in the case of seeing GOAWAY's or even other errors. But this isn't perfect and can still cause job failures to be seen for legitimate reasons when the Kube API isn't available.

The ask here is to make the current retry configurable with some variables. Unsure if retry time can be configured, but certainly number of retries being configurable would be nice to have.

Why does the customer need this? (List the business requirements here)

So that the amount of (un)expected failures seen due to issues communicating with the Kube API can be lessened.

How would you like to achieve this? (List the functional requirements here)

A setting for the Operator to ingest would be ideal.

Github Link

Original retry behavior was added here: https://github.com/ansible/receptor/pull/760