AnsweredAssumed Answered

Apache Drill System Error

Question asked by aalvarez on May 16, 2016
Latest reply on May 17, 2016 by Hao Zhu


Hi,

 

I am trying to join two existing views through Apache Drill but I am getting the error below and I am unable to find any documentation on how to solve it.

 

Here is the error message:

Error: SYSTEM ERROR: JaninoRuntimeException: Code attribute in class "org.apache.drill.exec.test.generated.HashJoinProbeGen4" grows beyond 64 KB

 

A generic version of the query (nothing too complicated):

0: jdbc:drill:> select * from `dfs.sf`.`view1` as t1

. . . . . . . > join

. . . . . . . > (select * from `dfs.sf`.`view2`) as t2

. . . . . . . > on t1.`variable1`=t2.`variable1`

. . . . . . . > ;

 

note that variable1 exists in both tables.

The view sizes are 33k and 31k. They both exist under the storage dfs.sf

 

The  memory configuration on drill-env.sh is:

DRILL_MAX_DIRECT_MEMORY="32G"

DRILL_HEAP="16G"

 

I appreciate if someone can help me figure out what could be the underlying problem.

 

Environment: drill 1.6 on Centos running  mapr 5.1.0.

Outcomes