Drizzle，一个精简版的MySQL分支，在目前的MySQL代码基本之上，将存储过程、视图、触发器、查询缓存、PREPARE语句等等没什么必要 的功能从代码中删掉，简化对数据类型和存储引擎的支持，并且进行大胆的重构。最终要实现的目的是将MySQL的代码大大简化，理顺MySQL的架构，改善 MySQL的代码质量，提高系统的稳定性和性能。将更适合Web应用、云计算组件、没有业务逻辑的数据库（又名存储过程）、多核架构等业务
这是 Drizzle 的第一个 GA 版本，该版本主要改进内容包括：
- Documentation is sphinx-based and available in-tree or online. It should be noted that they are licensed under Creative Commons!
- Replication solutions will be log based. Both file-based and innodb-table based versions of the log are available and have been heavily tested / validated.
- Native replication available via the slave plugin. This solution utilizes the innodb-based replication log and has been subjected to significant testing. Check out a simple example here. You can check out David Shrewsbury's replication related blog posts here
- Easy migration from MySQL systems via the Drizzledump tool.
- Drizzleadmin tool - used to bypass the protocol’s max-connections setting to do tasks such as killing queries or clients.
- libdrizzle is included in the tree (no longer a separate project) and is BSD licensed
- ANSI-compliant INFORMATION_SCHEMA
- DATA_DICTIONARY for custom, Drizzle-specific tables and data
- Innodb tables by default
- MyISAM available (temporary tables)
- Removed archive, blackhole, csv, and other engines (these can be reintroduced if faced with enough demand).
- HailDB engine = embedded Innodb, but better
- Pluggable authentication
- SQLAlchemy integration - we now pass their test suite and use it as part of our build and test system
- Significant performance improvements
- Initial work for catalogs
- And tons more!
发行说明：
http://blog.drizzle.org/2011/03/15/drizzle-2011-03-12-ga-tarball-has-been-released/
汗一个