Database engine development and common database internal principles Learning (1)

Database engine Development

The study of teacher Wang Tao's Database engine course has been a long time ago, limited to its own ability, especially the lack of understanding of the database, at that time can be said to swallow like learning, however, the harvest is still very large, and today, Think of their own understanding of the database to a higher level, but also because of the relatively complex and changeable modern database, the database model is much simpler than other databases, so re-turned out, again to learn, trying to clear the source layer of various database structure!

because it's a note-type blog post, it's constantly being updated.

Clue Database development relational database theory Introduction to NoSQL Databases

1. Record: Talking about the nature of database A database is a software that provides access to services Internet service available Provides high availability Provides high performance Provides a generalized interface to reduce learning costs across a variety of databases Why do I need a database A unified standard for database access to various language interfaces Provide reliable data Services Provides high-performance database access performance Data Management History The first thing to do is to save the data with a card. Tape Storage (1951) Disk Save (1956 IBM 5MB) Database management System Mesh Database (IDS) Hierarchical database (IBM's IMS) relational database (Oracle,db2,mysql) (1970 proposes a relational database theory base model, 1974 debates relational database, 1979 Oracle V2 release, the world's first commercially available database, 1983 support transactions, DB2 1981 years before the release of commercial, 1991 only to the common platform) Non-relational database (object model, XML storage, graph storage) Relational database theory is too demanding, it thrives on hardware performance improvements, inability to handle big data, and NoSQL is thriving. Database-dependent Technologies Operating system Store Internet Application Scenarios for Databases At the bottom level of the application The theoretical basis of the relational model All information is represented as data values in the relationship. Therefore, the relationship variables are not related to each other at design time; instead, the designer uses the same domain in multiple relationship variables, and if one property relies on another property, the dependency is enforced by referential integrity. NoSQL type Key-value Key-value pairs Achieve a leap in data by simplifying data operations Big-table Document documents The closest to a relational type of NoSQL The core lies in the nesting of relationships Graph graph SNS type of System Document-based Database model Solving the relational model is too idealistic Relational models are difficult to maintain Semi-structured for xml,json,rest services Easy and easy to learn Non-fixed field, non-fixed length data, non-stationary access 2. Simplification: The essence of a database is a software that provides data access services The database provides highly available, high-performance access services, and provides a unified development interface for each language platform Relational database is a very perfect mathematical theory of the database, flourished in the hardware performance of a large increase in hardware bottlenecks when the inability to process data beyond the scale. The cost of cross-server join operations is huge. As well as the theoretical perfection of the coupling degree of the data model, the model change is very difficult, the complexity of business change is very high NoSQL type Key-value Database BigTable Database Document Type Database Graphics database Document Type Database A NoSQL database closest to a relational database whose core is the relational join becomes nested With semi-structured data storage such as Xml,json, restful services can be conveniently provided Easy to learn, very flexible business model Non-fixed field, non-fixed length data, non-stationary access 3. Recite: The database is a highly available software that provides high-performance access services that unify the development interfaces of each language platform. In the distributed database, the main cost of the relational database is that the network cost of the cross-server join is very large, and the join itself is one of the most important characteristics of the relational database. A document database is a semi-structured database whose core is the relational join nesting, and because it stores semi-structured data such as Xml,json, it is very convenient to provide rest services. 4. Consider: Any merit comes at a price. There is no silver bullet, and one solution can solve all problems, especially under the constraints of hardware conditions The pros and cons of relational databases are a problem worth thinking about, only to solve this problem to really know when to use NoSQL, but this series of notes blog is mainly used to learn the principles of the database, this is not the focus, in the future may be issued a variety of NoSQL note posts. 5. Review:

Summarize Relational databases have limitations in the case of big data A document-based database is a semi-structured database

From for notes (Wiz)

Database engine development and common database internal principles Learning (1)