WeiWenda
WeiWenda
I draw a simple graph to illustrate what's happened when set spark.sql.catalogImplementation to xsql, or calling SparkSession.builder().enableXSQLSupport() 
As long as you change spark.sql.catalogImplementation to xsql, both spark-hive-thriftserver and spark-repl will start to support multi-datasource instead of one hive only. For permanent change, add new line `spark.sql.catalogImplementation=xsql` in...
@wushengyeyouya do you want to use hive thriftserver2 jdbc?
How does Scriptis&Linkis use spark now? xsql usage is all the same as spark
You can configure spark.sql.catalogImplementation=xsql in spark-defaults.conf,so that spark-shell、spark-sql、spark-submit will enable XSQL syntax. But above configuration only works in `bin` version, spark.sql.catalogImplementation=xsql is not allowed in raw spark. 
I new a project to show how to create and test new customized datasource manager. Please download and see https://github.com/WeiWenda/mock-datasource-manager.git
> @WeiWenda thanks for changing to `concurrentSQL` . > > It seems all tests fail with > > > No output has been received in the last 10m0s, this potentially...
> Hello, > > With this feature, would it mean that we can run several statement at the same time on one Livy session ? Regarding of course, that each...