Berkeley DB相关

为什么要使用Berkeley DB，它适合什么场合应用？
Berkeley DB并不适合所有的应用，因为简单，专一所以高效。
嵌入式数据库，的“嵌入”是指它内嵌在程序中，而不是说他只应用在嵌入式系统上。它的特点很适合应用于嵌入式系统上。当然在我们的pc机集群或大型服务器上，也可以灵活的配置，完成更艰巨的任务。

它适合于管理海量的，简单的数据。Google用Berkeley DB HA （High Availability） 来管理他们的帐户信息. Motorola在他的无线产品中用Berkeley DB跟踪移动单元。hp，microsoft，Sun Microsystems...等也都是它的大客户。它不能完全取代关系数据库，但在某些方面，它却有他们望尘莫及的高效性。

 

http://www.cnblogs.com/huqingyu/archive/2006/10/06/522251.html

 

http://www.ibm.com/developerworks/cn/linux/l-embdb/index.html

评论

1 楼 liudaoru 2010-09-27  

http://www.mail-archive.com/[e-mail]/msg53723.html

InnoDB scales *MUCH* better than BerkeleyDB.  While it did fine in our
benchmarks consisting of a small number of records, in some benchmarks we
ran here consisting of about a quarter million records, we let BerkeleyDB run
10 times longer than the InnoDB test before killing it and giving up.  The
live system is dealing with 2M records, and will soon be dealing with 2 to
3 times that many, so BerkeleyDB wasn't an option.

This system does quite a bit of inserting/updating, so for a read-mostly
table, the difference probably won't be as great, though from what I've
seen in some of our systems that use BerkeleyDB directly from C, there is
still some degradation from larger tables.

The only advantage of BerkeleyDB over InnoDB that I know of is that InnoDB
keeps everything in a set of files, whereas BerkeleyDB has a seperate file
per table.  Some people prefer it this way, as they can back up individual
files.  Personally, I prefer using mysqldump, which makes this a
non-issue for the most part.