Flink learning 3: data processing mode (stream batch)

Catalog ：

 

1. Data processing mode

At present, there are generally two types of data processing , Stream processing （ real-time processing , Generally for unbounded data sets ） And batch processing （ Offline processing , Generally, it deals with bounded data sets ）.

 

1.1 The batch , Generally, it deals with bounded data sets

1.2 Stream processing , Generally for unbounded datasets

 

 

 

 

There is a certain connection between bounded data sets and unbounded data sets , It is interchangeable under certain concepts .

 

therefore , There are some flow batch integrated frameworks for business , such as spark and flink

 

however , The flow batch integration principle of the two frameworks is different .