[HDFS-14864] DatanodeDescriptor Use Concurrent BlockingQueue

• Type: Improvement
• Status: Patch Available
• Priority: Minor
• Resolution: Unresolved
• Affects Version/s: 3.2.0
• Component/s: namenode
• Flags: Patch
• Assignee: David Mollitor
• Reporter: David Mollitor
• Created: 20/Sep/19 17:08
• Updated: 24/Sep/19 19:59

Description

https://github.com/apache/hadoop/blob/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/blockmanagement/DatanodeDescriptor.java#L104-L106

This collection needs to be thread safe and it needs to repeatedly poll the queue to drain it, so use BlockingQueue which has a drain() method just for this purpose:

This operation may be more efficient than repeatedly polling this queue.

https://docs.oracle.com/javase/7/docs/api/java/util/concurrent/BlockingQueue.html#drainTo(java.util.Collection,%20int)

Also, the collection returns 'null' if there is nothing to drain from the queue. This is a confusing and error-prone affect. It should just return an empty list. I've also updated the code to be more consistent and to return a java List in all places instead of a List in some and a native array in others. This will make the entire usage much more consistent and safe.