[HUDI-4205] Reading metadata table on S3 using Spark throws NullPointerException during createHFileReader

• Type: Bug
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 0.11.0
• Fix Version/s: 0.11.1
• Component/s: metadata, spark
• Labels: pull-request-available
• Story Points: 2
• Assignee: Ethan Guo (this is the old account; please use "yihua")
• Reporter: Ethan Guo (this is the old account; please use "yihua")
• Created: 07/Jun/22 22:35
• Updated: 21/May/24 19:55
• Resolved: 11/Jun/22 21:47

Description

Environment: EMR 6.6.0, OSS Spark 3.2.1, Hudi master

Storage: S3

When loading the metadata table in Spark shell using the following code, it throws NullPointerException.  In this case, the metadata table has the base files in HFile format.

This also happens for the following combinations: (1) Spark 3.1.3, Hudi 0.11.0 (1) Spark 3.2.1, Hudi 0.11.0 

spark.read.format("hudi").load("s3a://<base_path>/.hoodie/metadata/").show 

 

Caused by: java.lang.NullPointerException
  at org.apache.hudi.org.apache.hadoop.hbase.io.hfile.CacheConfig.<init>(CacheConfig.java:178)
  at org.apache.hudi.org.apache.hadoop.hbase.io.hfile.CacheConfig.<init>(CacheConfig.java:167)
  at org.apache.hudi.org.apache.hadoop.hbase.io.hfile.CacheConfig.<init>(CacheConfig.java:163)
  at org.apache.hudi.HoodieBaseRelation$.$anonfun$createHFileReader$1(HoodieBaseRelation.scala:531)
  at org.apache.hudi.HoodieBaseRelation.$anonfun$createBaseFileReader$1(HoodieBaseRelation.scala:482)
  at org.apache.hudi.HoodieMergeOnReadRDD.readBaseFile(HoodieMergeOnReadRDD.scala:130)
  at org.apache.hudi.HoodieMergeOnReadRDD.compute(HoodieMergeOnReadRDD.scala:100)
  at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:373)
  at org.apache.spark.rdd.RDD.iterator(RDD.scala:337)
  at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:52)
  at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:373)
  at org.apache.spark.rdd.RDD.iterator(RDD.scala:337)
  at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:52)
  at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:373)
  at org.apache.spark.rdd.RDD.iterator(RDD.scala:337)
  at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:90)
  at org.apache.spark.scheduler.Task.run(Task.scala:131)
  at org.apache.spark.executor.Executor$TaskRunner.$anonfun$run$3(Executor.scala:506)
  at org.apache.spark.util.Utils$.tryWithSafeFinally(Utils.scala:1462)
  at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:509)
  at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
  at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
  at java.lang.Thread.run(Thread.java:750) 

Spark shell:

./bin/spark-shell  \
     --master yarn \
     --deploy-mode client \
     --driver-memory 20g \
     --executor-memory 20g \
     --num-executors 2 \
     --executor-cores 8 \
     --conf spark.serializer=org.apache.spark.serializer.KryoSerializer \
     --conf spark.kryoserializer.buffer=256m \
     --conf spark.kryoserializer.buffer.max=1024m \
     --jars /home/hadoop/hudi-spark3.2-bundle_2.12-0.12.0-SNAPSHOT.jar \
     --conf 'spark.eventLog.enabled=true' --conf 'spark.eventLog.dir=hdfs:///var/log/spark/apps' \
     --conf 'spark.sql.extensions=org.apache.spark.sql.hudi.HoodieSparkSessionExtension' \
     --conf 'spark.sql.catalog.spark_catalog=org.apache.spark.sql.hudi.catalog.HoodieCatalog' 

 

 

Issue Links

• links to: GitHub Pull Request #5841