Fork PR upload should retry after GH rate limit when using tokenless upload.

[kaovilai]

The cli should wait and retry when using tokenless in fork PRs.

Code that logs when failing to upload

codecov-cli/codecov_cli/services/commit/__init__.py

Line 51 in d3f982f

logger.info("The PR is happening in a forked repo. Using tokenless upload.")

Example: vmware-tanzu/velero#7787

[thomasrockhu-codecov]

@kaovilai I believe this does try to retry after about 10 seconds. We cannot block CI on this rate limit as it can push CI builds into the hours of time.

More importantly, I think the underlying issue should be resolved with this issue