| Summary: | 碩士 === 中國科技大學 === 資訊工程系資訊科技應用碩士在職專班 === 105 === Relation database Through Key Value, describes the Relationship between Data Entity, Normalization to remove the contents of Consistency and Redundancy, resulting in Logical Database Schema (hereinafter referred to as Schema). So the database is designed to be Schema, not Record. Schema through the expansion, modification, and migration and other processing, the need for expensive maintenance costs. In addition to the data structure is not easy to change with the effectiveness of equipment alone for data processing.
Big data databases have many advantages over traditional relational databases. Storage costs are low, only need to increase the general computer equipment or expansion of the server node can achieve the same expansion of the same purpose. Data processing performance advantages, through the decentralized hardware architecture to spread data to multiple devices for data operations, take much less time than the traditional relational database. Data structure changes are easy; you can infinitely expand the field without the need to define the data type, which in the data structure changes will be easier. Based on the above reasons, the traditional relational database migration to the Column-Based large data database must be the future trend.
This study has the concept of system maintenance, from "top down" to the planning. The ER-model of the relational database is extended to EER-Model, and the schema will change as the data model changes.EE-R Model is object-oriented, and the data model of Column Based data database is also an object-oriented concept, and the rules of mutual conversion are found by similar data models. AS long as you can use this rule, you can convert the Schema of a traditional relational database to a Column Based database schema without the need for expensive maintenance costs.
|
|---|
