Get the App
SLTechnology News&Howtos  ›  Internet Technology  › 

Advantages and disadvantages of Hadoop Technology in big data's entry-level Learning

Shulou Source: shulou.com Published: 2022-06-02 19:16:55 11月16日 Update

Hadoop Technology Advantages and Disadvantages

(1)Hadoop has high reliability in terms of bit-wise storage and processing capabilities.

(2)Hadoop distributes data to complete storage and compute tasks through clusters of available computers, which can easily scale to thousands of nodes with high scalability.

(3)Hadoop can dynamically move data between nodes, and ensure the dynamic balance of each node, processing speed is very fast, with high efficiency.

(4)Hadoop can automatically save multiple copies of data and automatically redistribute failed tasks, with high fault tolerance.

. In the process of getting started learning big data, there are encounter learning, industry, lack of systematic learning route, systematic learning planning, welcome you to join my big data learning exchange skirt: 529867072, skirt file has my big data learning manual, development tools, PDF documents books, you can download by yourself.

Disadvantages of Hadoop

(1)Hadoop is not suitable for low-latency data access.

(2)Hadoop cannot efficiently store large numbers of small files.

(3)Hadoop does not support multiple users to write and modify files arbitrarily.

Core components of Hadoop

Since the birth of Hadoop, Hadoop1, Hadoop2 and Hadoop3 have appeared.

HDFS and MapReduce are the core components of Hadoop1, and many components in the Hadoop ecosystem are based on HDFS and MapReduce. Hadoop1 was followed by Hadoop2, which improved upon Hadoop1. Compared to Hadoop1, the three core components of Hadoop2 are HDFS, MapReduce and Yarn. Most businesses currently use Hadoop2, and this book uses Hadoop 2.7.3.

A common module and three core components of Hadoop2 make up four modules, which are described below.

Hadoop Common: Provides infrastructure for other Hadoop modules.

HDFS: Distributed file system with high reliability and high throughput.

(3)MapReduce: Based on Yarn system, distributed offline parallel computing framework.

(4)Yarn: A framework responsible for job scheduling and cluster resource management.

Tags: Data learning components files core system module node cluster storage three task distributed dynamic reliability foundation multiple framework version allocation Apple Docker Huawei Linux macOS MariaDB Microsoft MySQL NVidia OPPO Reno Redmi OPPO Reno macOS Shulou Information Microsoft