排名前10位的SQL和NoSQL数据库

jopen 10年前
数据库排名网站 DB Engines 用图表概述了目前近200种不同的数据库的排名情况。以下是排名前10位的SQL和NoSQL数据库。

The big three

Together the Oracle, MySQL and Microsoft heavyweight DBs take the lion’s share of users, each reeling in users with their own distinct advantages.

1. Oracle 11g

Initial release: 1980
License: Proprietary
SQL: Yes

  • The most popular choice for important commercial projects and the oldest major database in the market (not necessarily a disadvantage)

  • Available in four different editions: Enterprise, Standard, Standard Edition One, Express

  • More operating system flexibility than its Microsoft counterpart

  • The widest flexibility when it comes to the supported server operating systems

  • Virtual Private Database, Data Guard for standby database, automatic memory storage and undo management

2. MySQL

Initial release: 1995
License: Open source
SQL: Yes

  • Enterprises can start out using the free community server and later upgrade to the commercial version

  • Runs on Linux, Windows, OSX and FreeBSD and Solaris

  • Intuitive graphical user interface for designing database tables

  • Due to its open-source community, MySQL has a large bank of tutorials and information to help you get started and solve problems

  • Support for partitioning and replication, as well as for Xpath and for stored procedures, triggers and views

3. Microsoft SQL Server

Initial release: 1989
License: Proprietary
SQL: Yes

  • The most widely used commercial DBMS

  • Limited to Windows, but this is an advantage if your enterprise uses mostly Microsoft products

The alternatives

Over the years, Oracle, MySQL and Microsoft have firmly secured their place at the top of the DB market. But the DB-engines ranking depicts a steady increase in popularity among newer alternative databases, with services like HBase, Cassandra and Teradata slowly gaining ground.

排名前10位的SQL和NoSQL数据库

Source: db-engines.com

4. PostgreSQL

Initial release: 1989
License: Open source
SQL: Yes

  • A particularly scalable object-relational database

  • Runs on Linux, Windows, OSX and several other systems

  • Support for tablespaces, as well as for stored procedures, joins, views, triggers, etc.

5. MongoDB

Initial release: 2009
License: Open source
SQL: No

  • The most popular NoSQL DB; nevertheless retains some SQL properties like query and index

  • Supports a wide range of programming languages like Scala, Groovy, Clojure and Java – notably more than NoSQL rival Cassandra

  • High performance on large databases

  • Best for dynamic queries and for defining indexes

  • Supports Linux, OSX and Windows, but the DB size is limited to 2.5 GB on 32bit systems

6. DB2

Initial release: 1983
License: Proprietary
SQL: Yes and No

  • IBM’s answer to Oracle’s 11g, available in host and Windows/Linux versions

  • Runs on Linux, UNIX, Windows and mainframes

  • Ideal for IBM host environments

  • Support for both SQL and NoSQL data models

7. Microsoft Access

Initial release: 1992
License: Proprietary
SQL: Yes

  • Only one installation needed (DBMS and design tool in one)

  • Like Microsoft SQL Server, it’s use is limited Windows

  • Ideal for getting started with traffic analysis, but not its performance is not designed for mid to large-scale projects

  • Supported programming languages limited to C, C#, C++, Java, VBA and Visual Basic.NET

8. SQLite

Initial release: 2000
License: Open source
SQL: Yes

  • A self-contained server-less database engine with no external dependencies, used in iPhones, Firefox browsers and Skype

  • Its use as an embedded database makes it popular among devices like the iPhone, as well as desktop software like Skype and Firefox (SQLite even claims to be the most deployed SQL database)

  • No configuration or admin tasks needed

  • The entire DB is stored in one disk file which can grow to 7TB

  • The widest range of supported programming languages among the top 10 DBs

9. Cassandra

Initial release: 2008
License: Open source
SQL: No

  • Highly available NoSQL alternative to MongoDB

  • Useful for storing particularly large datasets with a user-friendly interface

  • Popular in banking, finance and logging, but also used by 非死book and 推ter  

  • Supports Windows, Linux and OSX, as well as numerous languages

  • Map/reduce also possible when used with Hadoop

10. Sybase ASE

Initial release: 1987
License: Proprietary
SQL: Yes

  • An enterprise RDBMS by SAP with

  • Runs on Linux, Unix and Windows, but no OSX support

  • Supported programming languages are limited to C, C++, Cobol, Java, Perl, PHP and Python

There are plenty more well-know DBs that didn’t make it into the top 10, for instance CouchDB (21), Neo4j (22) and Riak (30). Competition grows fierce further down the chain, with dozens more smaller and newer SQL and NoSQL databases for webmasters to choose from.

Check out the full ranking over on db-engines.com.

排名前10位的SQL和NoSQL数据库

Coman Hamilton