Real, real-time
Build applications that respond immediately to events. Craft materialized views over streams. Receive real-time push updates, or pull current state on demand.
Seamlessly leverage your existingApache Kafka®infrastructure to deploy stream-processing workloads and bring powerful new capabilities to your applications.
What, not how
Use a familiar, lightweight syntax to pack a powerful punch. Capture, process, and serve queries using only SQL. No other languages or services are required.
Build a complete streaming app with a few SQL statements.
Capture events
CREATE SOURCE CONNECTOR riders WITH ( 'connector.class'='JdbcSourceConnector', 'connection.url'='jdbc: postgresql: // ...', 'topic.prefix'='rider', 'table.whitelist'='geoEvents, profiles', 'key'='profile_id', ...);
Perform continuous transformations
CREATE STREAM locations AS SELECT rideId, latitude, longitude, GEO_DISTANCE (latitude, longitude, dstLatitude, dstLongitude AS kmToDst FROM geoEvents EMIT CHANGES;
Create materialized views
CREATE TABLE activePromotions AS SELECT rideId, qualifyPromotion (kmToDst) AS promotion FROM locations GROUP BY rideId EMIT CHANGES;
Serve lookups against materialized views
SELECT rideId, promotion FROM activePromotions WHERE ROWKEY='6fd0fcdb';
Simple constructs for building streaming apps.
ksqlDBenables you to build event streaming applications leveraging your familiarity with relational databases. Three categories are foundational to building an application:collections,stream processing, andqueries.
Streams
Streamsare immutable, append -only sequences of events. They’re useful for representing a series of historical facts.
CREATE STREAM routeWaypoints ( vehicleId VARCHAR, latitude DOUBLE (10, 2), longitude DOUBLE (10, 2) (WITH) kafka_topic='locations', partitions=3, key='vehicleId', value_format='json' );
Tables
Tablesare mutable collections of events. They let you represent the latest version of each value per key.
CREATE TABLE currentCarLocations ( vehicleId VARCHAR, latitude DOUBLE (10, 2), longitude DOUBLE (10, 2) (WITH) kafka_topic='locations', partitions=3, key='vehicleId', value_format='json' );
Stream processingenables you to execute continuous computations over unbounded streams of events, ad infinitum. Transform, filter, aggregate, and join collections together to derive new collections or materialized views that are incrementally updated in real-time as new events arrive.
Push
Push querieslet you subscribe to a query’s result as it changes in real-time. When new events arrive, push queries emit refinements, which allow you to quickly react to new information. They’re a perfect fit for asynchronous application flows.
SELECT vehicleId, latitude, longitude FROM currentCarLocations WHERE ROWKEY='6fd0fcdb' EMIT CHANGES;
Pull
Pull queriesallow you to fetch the current state of a materialized view. Because materialized views are incrementally updated as new events arrive, pull queries run with predictably low latency. They’re a great match for request / response flows.
SELECT vehicleId, latitude, longitude FROM currentCarLocations WHERE ROWKEY='6fd0fcdb';
One mental model for the entire stack.
Today, nearly all streaming architectures are complex, piecemeal solutions. They comprise multiple subsystems, each with its own mental model.
With a lightweight, familiar SQL syntax,ksqlDB presents a single mental modelfor working with event streams across your entire stack: event capture, continuous event transformations, aggregations, and serving materialized views.
GIPHY App Key not set. Please check settings