Re: [DISCUSS] Preventing Mockito usage for the new code with Checkstyle

Hey Jing, That's basically up to us to decide -- maybe a continuous approach is safer? (tests with junit4 behaviour won't be mergeable if we enable rules immediately) That said, looks like the migration work is almost there -- with some effort we

Re: [DISCUSS] Preventing Mockito usage for the new code with Checkstyle

Thanks @Panagiotis for the hint! Does it mean that those suppressions need to be cleaned up continuously while we move forward with Junit5 migration(extra guideline is required), even if regex has been used? Or we just leave them as they are and clean them up in one shot after every junit4 has been

Re: [DISCUSS] FLIP 295: Support persistence of Catalog configuration and asynchronous registration

Hi Feng, There are still many places contain inconsistent content, e.g. 1. "Asynchronous registration" is still used. 2. The java comment of the method registerCatalog(String catalogName, Catalog catalog) in CatalogManager does not tell what the method is doing. There might be more such issues.

[jira] [Created] (FLINK-31977) If scaling.effectiveness.detection.enabled is false, the call to the detectIneffectiveScaleUp() function is unnecessary

Tan Kim created FLINK-31977: --- Summary: If scaling.effectiveness.detection.enabled is false, the call to the detectIneffectiveScaleUp() function is unnecessary Key: FLINK-31977 URL: https://issues.apache.org/jira/browse/

[jira] [Created] (FLINK-31976) Once marked as an inefficient scale-up, further scaling may not happen forever

Tan Kim created FLINK-31976: --- Summary: Once marked as an inefficient scale-up, further scaling may not happen forever Key: FLINK-31976 URL: https://issues.apache.org/jira/browse/FLINK-31976 Project: Flink