guanziyue removed a comment on pull request #4913: URL: https://github.com/apache/hudi/pull/4913#issuecomment-1065343231
@nsivabalan By the way, the following case may be handled in a more efficient way if we have this PR. Looking forward to your opinion. In one commit, task attempt 1 create a log file .fileId_instantTime.log.2_0-0-1. But this task attempt failed. task attempt 2 create a new log file .fileId_instantTime.log.2_0-0-2. This task attempt succeeds and .fileId_instantTime.log.2_0-0-2 is returned in writeStat. Before: log file is fail-safe. So we preserve both of them and it should be idempotent to handle the content twice. Now: we can delete .fileId_instantTime.log.2_0-0-1 in 'reconcileAgainstMarkers' phase because it is an invalid file, which improves perf little. But this way no longer treat log file fail-safe totally. -- 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]
