Hadoop-the Definitive Guide, 4th Edition (learning side translation) Chapter 1

Chapter 1. Meet Hadoop (mostly slightly)

Comparison with other Systems

Talk about comparisons with other systems.

Relational Database Management Systems

Why don't we use a lot of hard drives for large-scale analysis? Why Hadoop is needed?

The answer is: In all these years, disk seek time has increased much more slowly than the transfer rate. The Seek is to move the head to the specified place for reading. Seek time is a physical operation, so there is a delay. However, the transfer rate depends on the bandwidth of the hard disk.

On the one hand, if the pattern of data reading is limited by seek, it will take more time to read and write most of the data. On the other hand, when a large amount of data is updated, the traditional database b-tree is not as effective as mapreduce, and MapReduce uses Sort/merge to reconstruct the database.

Hadoop-the Definitive Guide, 4th Edition (learning side translation) Chapter 1