04 - Apache Spark for Java Developers - Local vs Distributed Systems смотреть онлайн

@backstreetbrogrammer

--------------------------------------------------------------------------------
Chapter 01 - Apache Spark for Java Developers - Local vs Distributed Systems
--------------------------------------------------------------------------------
Big data can not be processed or stored in a local system or a single node. It requires multiple machines or nodes to store / process it.

Master Node =: Worker nodes

A local single node will use the computation sources (CPU, cores) and storage (memory, hard disk) of a single machine only. Only vertical scaling is possible which means we can add powerful CPU or memory to a single machine but there will be a limit to it. Single point of failure if the local node goes down which makes it essential to store the important data in cloud or separate disk.

A distributed system has access to the computation sources (CPU, cores) and storage (memory, hard disk) across a number of machines connected through a network. Horizontal scaling is easier by just adding new nodes or systems to the distributed system. It also supports fault tolerance, if one machine fails, the whole network can still go on.

Github: https://github.com/backstreetbrogrammer/11_JavaSpark

- Apache Spark for Java Developers Playlist: https://youtube.com/playlist?list=PLQDzPczdXrTgqEc0uomGYDS0SFu7qY3g3
- Java Serialization Playlist: https://youtube.com/playlist?list=PLQDzPczdXrTjGLkI8kd2d0bfNHx0v7dTy
- Dynamic Programming Playlist: https://youtube.com/playlist?list=PLQDzPczdXrTilRgmNUQ632OrD-wfRLUR_

#java #javadevelopers #javaprogramming #apachespark #spark