SEZ9 commented on issue #11911: URL: https://github.com/apache/seatunnel/issues/11911#issuecomment-5366245873
Well-scoped follow-up from the #11827 review. Agreed that the retry should live inside flush() so it applies uniformly to all callers (batch_size, timer flush, prepareCommit, close) rather than only the checkpoint path. Please check whether an existing retry utility in the codebase (e.g., common retry helpers used by other connectors such as RetryUtils) can be reused instead of adding a new implementation. Also consider which HTTP status codes should be retried (e.g., 5xx/429 and transport errors) versus failed fast (4xx client errors). Suggest exposing max attempts and backoff (initial + cap) as sink options with conservative defaults, keeping retries bounded so a genuine outage still fails loudly. New options should be documented in the Prometheus sink docs and covered by a test simulating a transient failure. <!-- streview-comment:426 --> -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
