flink-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Fabian Hueske <fhue...@gmail.com>
Subject Design Doc: Execution of Table API & SQL queries in Java/Scala programs
Date Wed, 09 Mar 2016 15:52:21 GMT
Hi everybody,

the following design document shows different ways to execute Table API and
SQL queries (query in the following) in Java and Scala programs. It
proposes alternatives for data sources and result sinks of queries.

Data Sources
- Use a DataSet, DataStream, or Table as input for a query
- Use an external table (CSV, Parquet, ORC, Hbase, …) as input for a query
- Register external catalogs and use their tables as input for a query

Result Sinks
- Emit the result of a query as DataSet, DataStream, or Table
- Emit the result to an external storage system (file, JDBC, Kafka, …)

These options allow to embed a query into a regular DataSet / DataStream
program, as well as, specifying it without touching the DataSet or
DataStream API at all by reading from external tables and writing to an
external system.

At the moment, Table API queries can only use DataSets as input and emit
their results as DataSets. SQL queries are not supported yet.


Feedback and comments are welcome.

Cheers, Fabian

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message