[HDFS-11879] Fix JN sync interval in case of exception

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 3.0.0-alpha4
• Hadoop Flags: Reviewed
• Assignee: Hanisha Koneru
• Reporter: Hanisha Koneru
• Created: 24/May/17 19:21
• Updated: 24/Apr/18 20:51
• Resolved: 25/May/17 21:03

Description

Currently, if there is an exception in JournalNodeSyncer#syncJournal, the sync daemon does not wait the specified interval before the next sync attempt.

In the following code block in JournalNodeSyncer#startSyncJournalsDaemon, Thread.sleep should be moved to a separate try catch block from syncJournal() to avoid skipping it in case of exception.

        try {
          if (!journal.isFormatted()) {
            LOG.warn("Journal cannot sync. Not formatted.");
          } else {
            syncJournals();
          }
          Thread.sleep(journalSyncInterval);
        } catch (Throwable t) {
           .......