If the main instance is unavailable to accept the build result, the worker should retry a few times, giving some time in between for the instance to come back.
At the moment the worker is just going to discard the result and return to the regular "request more work until task was assigned" loop.
If the main instance is unavailable to accept the build result, the worker should retry a few times, giving some time in between for the instance to come back.
At the moment the worker is just going to discard the result and return to the regular "request more work until task was assigned" loop.