[HDFS-119] logSync() may block NameNode forever.

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.21.0, 1.0.3
• Component/s: namenode
• Hadoop Flags: Reviewed
• Assignee: Suresh Srinivas
• Reporter: Konstantin Shvachko
• Created: 16/Mar/09 23:54
• Updated: 24/Apr/12 21:54
• Resolved: 30/Jul/09 22:36

Description

1. FSEditLog.logSync() first waits until isSyncRunning is false and then performs syncing to file streams by calling EditLogOutputStream.flush().
   If an exception is thrown after isSyncRunning is set to true all threads will always wait on this condition.
   An IOException may be thrown by EditLogOutputStream.setReadyToFlush() or a RuntimeException may be thrown by EditLogOutputStream.flush() or by processIOError().
2. The loop that calls eStream.flush() for multiple EditLogOutputStream-s is not synchronized, which means that another thread may encounter an error and modify editStreams by say calling processIOError(). Then the iterating process in logSync() will break with IndexOutOfBoundException.

Issue Links

• is related to: HADOOP-4045 Increment checkpoint if we see failures in rollEdits
• relates to: HDFS-192 TestBackupNode sometimes fails