[HIVE-24862] Fix race condition causing NPE during dynamic partition loading

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 4.0.0-alpha-1
• Labels: pull-request-available
• Assignee: Zoltan Chovan
• Reporter: Rajesh Balamohan
• Created: 09/Mar/21 04:00
• Updated: 17/Nov/22 08:48
• Resolved: 11/Mar/21 09:54

Description

Following properties default to 15 threads.

hive.load.dynamic.partitions.thread
hive.mv.files.thread  

During loadDynamicPartitions, it ends ups initializing newFiles without synchronization (HIVE-20661, HIVE-24738).
https://github.com/apache/hive/blob/master/ql/src/java/org/apache/hadoop/hive/ql/metadata/Hive.java#L2871

This causes race condition when dynamic partition thread internally makes use of hive.mv.files.threads in copyFiles/replaceFiles.
This causes "NPE" during retrieval in addInsertFileInformation().

 

e.g stacktrace

Caused by: java.lang.NullPointerException
  at org.apache.hadoop.fs.FileSystem.fixRelativePart(FileSystem.java:2734)
  at org.apache.hadoop.hdfs.DistributedFileSystem.fixRelativePart(DistributedFileSystem.java:3396)
  at org.apache.hadoop.hdfs.DistributedFileSystem.getFileStatus(DistributedFileSystem.java:1740)
  at org.apache.hadoop.fs.FileSystem.isDirectory(FileSystem.java:1740)
  at org.apache.hadoop.hive.ql.metadata.Hive.addInsertFileInformation(Hive.java:3566)
  at org.apache.hadoop.hive.ql.metadata.Hive.fireInsertEvent(Hive.java:3540)
  at org.apache.hadoop.hive.ql.metadata.Hive.loadPartitionInternal(Hive.java:2414)
  at org.apache.hadoop.hive.ql.metadata.Hive.lambda$loadDynamicPartitions$4(Hive.java:2909)
  at java.util.concurrent.FutureTask.run(FutureTask.java:266)
  at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
  at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
  at java.lang.Thread.run(Thread.java:748)

Issue Links

• links to: GitHub Pull Request #2053