You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@spark.apache.org by wuchang <58...@qq.com> on 2017/06/17 09:27:38 UTC

Build spark without hive issue, spark-sql doesn't work.

I want to build hive and spark to make my hive work on spark engine.
I choose Hive 2.3.0 and Spark 2.0.0, which is claimed to be compatible by hive official document.
According to the hive officials document ,I  have to build spark without hive profile to avoid the conflict between original hive and spark-integrated hive. 
Yes, I build successfully , but then the problem comes:I cannot use spark-sql anymore because spark-sql relies on the hive library and my spark is a no-hive build.

I don’t know the relationship between hive-integrated hive and original hive. 
Below is the spark-integrated hive jars:

hive-beeline-1.2.1.spark2.jar
hive-cli-1.2.1.spark2.jar
hive-exec-1.2.1.spark2.jar
hive-jdbc-1.2.1.spark2.jar
hive-metastore-1.2.1.spark2.jar
spark-hive_2.11-2.0.0.jar
spark-hive-thriftserver_2.11-2.0.0.jar

It seems that Spark 2.0.0 relies on hive 1.2.1.
How can I build and setup spark and to make hive on spark
Work properly and my spark-sql、pyspark and spark-shell work properly?