[GitHub] [hbase] Apache9 commented on pull request #4038: HBASE-26552 Introduce retry to logroller when encounters IOException

2022-01-26 Thread GitBox
Apache9 commented on pull request #4038: URL: https://github.com/apache/hbase/pull/4038#issuecomment-1022296101 OK, I got your point. But two levels of retry will greatly increase the duration of retries... If this is the case, maybe we should move the retry to the upper layer only?

[GitHub] [hbase] Apache9 commented on pull request #4038: HBASE-26552 Introduce retry to logroller when encounters IOException

2022-01-23 Thread GitBox
Apache9 commented on pull request #4038: URL: https://github.com/apache/hbase/pull/4038#issuecomment-1019479511 I think we already have retries when creating the writer? See FanOutOneBlockAsyncDFSOutputHelper.createOutput. -- This is an automated message from the Apache Git Service. To