Chinaunix首页 | 论坛 | 博客
  • 博客访问: 174699
  • 博文数量: 46
  • 博客积分: 1445
  • 博客等级: 上尉
  • 技术积分: 448
  • 用 户 组: 普通用户
  • 注册时间: 2008-05-27 22:20
文章分类

全部博文(46)

文章存档

2013年(4)

2012年(12)

2011年(3)

2010年(5)

2009年(16)

2008年(6)

我的朋友

分类: Java

2012-12-13 05:51:42

Sphinx is a modified mysql server optimized for searching. It is reasonably fast, but scaling it is DIY.

Solr is a well supported Lucene based search engine. It has one of the most advanced feature sets amongst search engines out there and is very stable with great community. 

ElasticSearch is a relative new addition in search field, Lucene based search engine. Its "Elastic" in a sense that scaling it to multiple machines is built-in and very easy. It supports sharding and/or replication. This is an engine of choice for me, as it is trivial to setup, it has a great JSON API for doing all the operations and it is very customizable.

CloudSearch is a proprietary creation of Amazon. It's a paid setup and amount of money you pay depends on amount of traffic and requests that it serves. Ultimately, it's not a good fit for me, since it lacks geolocation search ( by lat/long ). It's hard to determine if it's Lucene based, but it's certainly not open-source, so if you have an issue, or you want to modify something in the search behaviour, which is not supported by Amazon currently - you are out of luck ( pretty much like every other closed source solution ). Also a word of warning, we've noticed with a number of Amazon services that their stated performance is actually lower in real life. So at times you have to provision for a higher usage than what you've anticipated that you have to pay for.

阅读(771) | 评论(0) | 转发(0) |
0

上一篇:sharding方式

下一篇:lucene阅读笔记

给主人留下些什么吧!~~