Package org.apache.flink.streaming.examples.dsv2.watermark

Class CountSales

java.lang.Object

org.apache.flink.streaming.examples.dsv2.watermark.CountSales

public class CountSales
extends Object

This example shows how to count the cumulative sales of each product at the moment. In this example, we simulate the creation and propagation of event time watermarks through Watermark.

Please note that Flink provides users with an event time extension to support event time. You can refer to the CountNewsClicks example.

The example uses a DataGeneratorSource as input source and a custom Watermark to propagate the event time in the stream.

Usage:

• --parallelism <path>The parallelism of the source. The default value is 5.

This example shows how to:

• How to define and propagate custom event (Watermark) in DataStream API V2

Please note that if you intend to run this example in an IDE, you must first add the following VM options: "--add-opens=java.base/java.util=ALL-UNNAMED". This is necessary because the module system in JDK 17+ restricts some reflection operations.

Please note that the DataStream API V2 is a new set of APIs, to gradually replace the original DataStream API. It is currently in the experimental stage and is not fully available for production.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	CountSales.CumulativeSales	CountSales.CumulativeSales represents the cumulative sales at a certain moment of a product.
static class	CountSales.Order	Pojo class for Order.

Field Summary

Fields

Modifier and Type	Field	Description
static final org.apache.flink.api.common.watermark.LongWatermarkDeclaration	EVENT_TIME_WATERMARK_DECLARATION	Firstly, we define an event time watermark, which represents the time of currently processing event.

Constructor Summary

Constructors

Constructor	Description
CountSales()

Method Summary

Modifier and Type	Method	Description
static void	main(String[] args)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

EVENT_TIME_WATERMARK_DECLARATION

public static final org.apache.flink.api.common.watermark.LongWatermarkDeclaration EVENT_TIME_WATERMARK_DECLARATION

Firstly, we define an event time watermark, which represents the time of currently processing event. Since the watermark needs to convey the timestamp, its data type is long. To determine the minimum event time across all watermarks, we utilize the combineFunctionMin() method to combine the watermarks. The default handling strategy is forward, meaning that the watermark will typically be advanced to downstream operators in most scenarios. Thus, we create a WatermarkDeclaration instance that can be used to declare and generate the watermark.

Constructor Details

CountSales

public CountSales()

Method Details

main

public static void main(String[] args)
                 throws Exception

Throws:

Exception