Stream

If you have experienced relational databases, you should know that tables are the core data models, which define types and constraints of columns.

Streams in SpringQL are very similar to tables. They have columns, and columns have types and constraints. Unlike tables, however, streams do not persist rows. Rows flow through streams and then go to downstream streams (often just called downstreams).

Here is an example stream definition in SQL:

Stream definition in SQL

CREATE STREAM s (
    ts TIMESTAMP NOT NULL ROWTIME,    
    c1 INTEGER NOT NULL
);

This statements defines stream s with two columns: ts and c1. ts is a TIMESTAMP type and should not take NULL value. c1 is an INTEGER type and also should not take NULL value.

ts has ROWTIME keyword, which means the column works as an event time (see Event time and watermark) for rows going through this stream.

See the Data type page for more information about columns' definition.

Rowtime, event time, and processing time

Every row in SpringQL has a rowtime. A rowtime is either a:

• Event time: when the stream definition includes a ROWTIME keyword
• Processing time: otherwise.
  • Processing time is the time when the row is inserted to the stream.

Source streams and sink streams

SpringQL provides not only "normal" streams, but also source streams and sink streams.

Source streams are streams to input foreign data via source readers, and sink streams are ones to output foreign data via sink writers.

The 3 types of streams are differently defined in SpringQL but currently there are no functional differences.