gitedmond opened a new pull request, #23456: URL: https://github.com/apache/kafka/pull/23456
### Summary The tiered-storage task retry settings were defined but never used. Every task was scheduled with the regular task interval, so retryable remote-storage failures did not honor the configured backoff. This change reschedules copy, expiration, and follower tasks one run at a time. Retryable failures use the existing exponential backoff settings; successful, skipped, and non-retryable runs return to the normal interval. Cancellation is synchronized with replacement-future publication so a cancelled task cannot re-enqueue itself. The contribution is my original work and is licensed to the Apache Software Foundation under the Apache License, Version 2.0. ### Testing - `./gradlew :storage:test --tests org.apache.kafka.server.log.remote.storage.RemoteLogManagerTest` - `./gradlew :storage:checkstyleMain :storage:checkstyleTest :storage:spotlessCheck` - `git diff --check` The targeted suite passes, including deterministic retry progression, jitter/cap behavior, reset behavior, leadership cancellation, self-cancellation, and cancellation during rescheduling. The full `:storage:test` run was attempted but is blocked by Windows file-lock failures in existing `LogLoaderTest` and `AbstractIndexTest` cases; the same failure reproduces on unchanged trunk. -- 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]
