[jira] [Resolved] (HBASE-30255) Async archiving wal file causes TestLogRolling flaky

["Duo Zhang (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HBASE-30255?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Duo Zhang resolved HBASE-30255.
-------------------------------
    Fix Version/s: 2.7.0
                   3.0.0-beta-2
                   2.5.16
                   2.6.7
     Hadoop Flags: Reviewed
       Resolution: Fixed

Pushd to all active branches.

Thanks [~liuxiaocs] and [~paksyd] for reviewing!

> Async archiving wal file causes TestLogRolling flaky
> ----------------------------------------------------
>
>                 Key: HBASE-30255
>                 URL: https://issues.apache.org/jira/browse/HBASE-30255
>             Project: HBase
>          Issue Type: Sub-task
>          Components: test, wal
>            Reporter: Duo Zhang
>            Assignee: Duo Zhang
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 2.7.0, 3.0.0-beta-2, 2.5.16, 2.6.7
>
>
> {code}
>     if (logsToArchive != null) {
>       final List<Pair<Path, Long>> localLogsToArchive = logsToArchive;
>       // make it async
>       for (Pair<Path, Long> log : localLogsToArchive) {
>         logArchiveExecutor.execute(() -> {
>           archive(log);
>         });
>         this.walFile2Props.remove(log.getFirst());
>       }
>     }
> {code}
> Here we will use a logArchiveExecutor to archive the wal file, and in the archive method we will decrease the log file size, but the removal from walFile2Props is done in foreground, so the assertion in TestLogRolling is flaky
> {code}
>     // should have deleted all the rolled wal files
>     TEST_UTIL.waitFor(5000, () -> AbstractFSWALProvider.getNumRolledLogFiles(log) == 0);
>     assertEquals(0, AbstractFSWALProvider.getLogFileSize(log));
> {code}
> As when num rolled fles reaches zero, the log file size may not be decreased yet.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)