It allows for for both batch and streaming analytics

It relies on relational databases for quick retrieval of data

It eliminates the need for data replication in distributed systems

It runs MapReduce for stream processing of real-time data

It spins out AWS Lambda functions to process data

It provides high availability for pub/sub topics and real-time data streams.   

It runs machine learning algorithms using Kafka user-defined functions.   

It implements a Lambda architecture for batch and streaming processing of factory sensor data.   

It facilitates data replication from the core cloud to the edge.   

MetadataNode

NodeManager

NameNode

ResourceManager

DataNode