Originally released in the third border of the Institute szhshp , please indicate
Selected because it is the easy ah
Relational Database
Feature
- Based on a single relational model, structured storage, there are integrity constraints
- Establish a link between the data by two-dimensional table
- Using Structured Query Language (SQL) for data reading and writing
- Saving data consistency
advantage
- Maintain data consistency through transaction processing
- Data update overhead is small
- Join and so on can be complex queries
- Technology is mature
Shortcoming
- Sql data reader must parse large amounts of data, lack of read and write performance under high concurrency
- To ensure data consistency, you need to lock the impact concurrent operations
- Unable to adapt to unstructured storage
- Generally large burden on the server
- Sometimes the data is too abstract, that is, objects stored in the database with the actual target entities have some differences
- Extended difficult
- Database huge, expensive
Non-relational databases
Feature
- Unstructured storage
- Based on the multi-dimensional relational model
- Easy deployment, open source free, low cost
Advantage
- High concurrency processing, ability of large quantities of data
- Support distributed clustering, load balancing, high performance
- Memory-level database, query speed
- Multi-storage format that supports key-value form, the form of documents, images form
no limit multi-table join query mechanisms, high scalability
Shortcoming
- Technical late start, limited maintenance tools and technical information
- It does not support the industry standard sql
- Not join other complicated connection operation
- Transaction Capabilities weak
- No integrity constraints, support for complex business scenarios poor