Key-Value数据库 Flare

     <p><a href="/misc/goto?guid=4959498018442349495" rel="nofollow">Tokyo Cabinet</a>(TC) 是日本第一大SNS网站mixi开发的，而Flare是日本第二大<span style="font-family:'Calibri', 'sans-serif';" lang="EN-US">SNS</span>网站green.jp开发的。Flare简单的说就是给TC添加了scale功能。他替换掉了TT部分，自己另外给TC写了网络服务器，Flare的主要特点就是支持scale能力，他在网络服务端之前添加了一个node server，来管理后端的多个服务器节点，因此可以动态添加数据库服务节点，删除服务器节点，也支持failover。如果你的使用场景必须要让TC可以scale，那么可以考虑flare。</p>    <p>flare唯一的缺点就是他只支持<a href="/misc/goto?guid=4959498018552551318" rel="nofollow">memcached</a>协议，因此当你使用flare的时候，就不能使用TC的table数据结构了，只能使用TC的key-value数据结构存储。</p>    <p>Flare 是分布式，可持久化的key-value存储系统，它具有以下特性：</p>    <ul style="padding-left:16px;margin-left:16px;" class="list1">     <li>persistent storage (you can use flare as persistent memcached)</li>     <li>pluggable storage (currently only <a href="/misc/goto?guid=4959498018442349495" rel="nofollow">Tokyo Cabinet</a> is available, though:)</li>     <li>data replication (synchronous or asynchronous)</li>     <li>data partitioning (automatically partitioned according to # of master servers (clients do not have to care about it))</li>     <li>dynamic reconstruction, and partitioning (you can dynamically (I mean, without any service interruption) add slave servers and partition master servers)</li>     <li>node monitoring and failover (if any server is down, the server is automatically isolated from active servers and another slave server is promoted to master server)</li>     <li>request proxy (you can always get same result regardless of servers you connect to. so you can think flare servers as one big key-value storage)</li>     <li>over 256 bytes keys, and over 1M bytes values are available</li>    </ul>    <p><strong>项目主页：</strong><a href="http://www.open-open.com/lib/view/home/1322721807874" target="_blank">http://www.open-open.com/lib/view/home/1322721807874</a></p>