Hi: The steps to compile CarbonData 1.5.1 with Spark 2.3.1 are as follows: 1. cover CarbonDataSourceScan.scala: cp -f integration/spark2/src/main/commonTo2.1And2.2/org/apache/spark/sql/execution/strategy/CarbonDataSourceScan.scala integration/spark2/src/main/spark2.3/org/apache/spark/sql/execution/strategy/CarbonDataSourceScan.scala
2. Edit integration/spark-common-test/src/test/scala/org/apache/carbondata/spark/testsuite/bigdecimal/TestBigDecimal.scala: Line 48: change 'salary decimal(30, 10))' to 'salary decimal(27, 10))' 3. Edit integration/spark-common/src/main/scala/org/apache/spark/util/CarbonReflectionUtils.scala: 1) Line 297: change 'classOf[Seq[String]],' to 'classOf[Seq[Attribute]],' 2) Replace Line 299-301 with a line 'method.invoke(dataSourceObj, mode, query, query.output, physicalPlan)'; 4. Use command to compile: mvn -DskipTests -Pspark-2.3 -Phadoop-2.8 -Pbuild-with-format -Pmv -Dspark.version=2.3.1 -Dhadoop.version=2.6.0-cdh5.8.3 clean package. It worked. You can refer to the PR#2779 <https://github.com/apache/carbondata/pull/2779> . -- Sent from: http://apache-carbondata-dev-mailing-list-archive.1130556.n5.nabble.com/