[Solved] Spark job failed during runtime. Please check stacktrace for the root cause.

hive on spark reports an error
executing the hive command is an error

[42000][3] Error while processing statement: FAILED: Execution Error, return code 3 from org.apache.hadoop.hive.ql.exec.spark.SparkTask. Spark job failed during runtime. Please check stacktrace for the root cause.

[Reason]
View running tasks on yarn, Query error results from error log

Map operator initialization failed: org.apache.hadoop.hive.ql.metadata.HiveException: Unexpected column vector type LIST

List type error
List corresponds to array in hive, array corresponds to list in Java

[Solution]
Temporarily change the execution engine to MR

set hive.execution.engine=mr;

There are many bugs in hive on spark, When an unknown error occurs, First try to replace the underlying execution engine with MR, to execute the sql statement.

[Subsequent modification]
1. View the current execution engine of hive:

set hive.execution.engine;

2. Manually set hive’s current execution engine to Spark

set hive.execution.engine=spark;

3. Manually set hive’s current execution engine to MR

set hive.execution.engine=mr;

Read More:

Leave a Reply

Your email address will not be published. Required fields are marked *