- When extracting HBase namespace, it returns ref instead of real namespace.
- When using ET JDBC to update MySQL, Losing data will happens.
- more......
- Python
- Batch
- Stream
- Remove streamingpro-opencv, streamingpro-dl4j, streamingpro-automl
- Merge streamingpro-crawler into stremingpro-mlsql
/stream/jobs/killand/stream/jobs/running,/run/sqlare removed
- PyJava is an ongoing effort towards bringing the data exchanging ability between Java/Scala and Python.
- MLSQL Console supports pyechars render.
- MLSQL Console supports notebook mode
- Python support should be more robust
- Remove Spark 2.2.x support
- Dynamically increase/decrease engine resource
- Table cache
- sessionPerRequest create new session for every request.
- directQuery Auth support directQuery auth.
- set statement sql auth
- Stream callback
- Stream batch mode(for sink)
- MySQL Binlog datasource
- Delta plus
- Stream schema infer
- Kafka tool
- Script progress track you can use "load mlsql.
/jobs/get/[jobid]as output" to get the script processing.
- Adding new symbol
!to execute command in MLSQL. - Grammar validate before really execute MLSQL script.
- PR-1011 Jython udf do not support null parameter
- PR-1010 Kill cannot kill job when sessionPerUser enabled
- PR-1000 System show jobs without stream jobs.
- Use streamJDBC instead of JDBC in stream sink.
- Use streamParquet instead of parquet in stream sink.
- MLSQL Console can be used with MLSQL Engine.
- API
/stream/jobs/killand/stream/jobs/running,/run/sqlwill be removed in next release. - Module
streamingpro-automlwill be removed in next release.