Known Issues in Apache Spark

This topic describes known issues and workarounds for using Spark in this release of Cloudera Runtime.

CDPD-217: HBase/Spark connectors are not supported

The Apache HBase Spark Connector (hbase-connectors/spark) and the Apache Spark - Apache HBase Connector (shc) are not supported in the initial CDP release.

Workaround: None

CDPD-3038: Launching pyspark displays several HiveConf warning messages

When pyspark starts, several Hive configuration warning messages are displayed, similar to the following:

19/08/09 11:48:04 WARN conf.HiveConf: HiveConf of name hive.vectorized.use.checked.expressions does not exist
19/08/09 11:48:04 WARN conf.HiveConf: HiveConf of name hive.tez.cartesian-product.enabled does not exist

Workaround: These errors can be safely ignored.

CDPD-2650: Spark cannot write ZSTD and LZ4 compressed Parquet to dynamically partitioned tables

Workaround: Use a different compression algorithm.

CDPD-3293: Cannot create views (CREATE VIEW statement) from Spark

Apache Ranger in CDP disallows Spark users from running CREATE VIEW statements.

Workaround: Create the view using Hive or Impala.

CDPD-3783: Cannot create databases from Spark

Attempting to create a database using Spark results in an error similar to the following:

org.apache.spark.sql.AnalysisException:
            org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:Permission denied: user [sparkuser] does not have [ALL] privilege on [hdfs://ip-10-1-2-3.cloudera.site:8020/tmp/spark/warehouse/spark_database.db]);

Workaround: Create the database using Hive or Impala, or specify the external data warehouse location in the create command. For example:

sql("create database spark_database location '/warehouse/tablespace/external/hive/spark_database.db'")