# 集群查询变慢

**URL:** <https://discuss.elastic.co/t/topic/110732>\
**Category:** 中文提问与讨论\
**Created:** [December 7, 2017, 6:33pm UTC](https://discuss.elastic.co/t/topic/110732 "2017-12-07T18:33:36Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![Zhengcong\_Yin](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/zhengcong_yin/32/62960_2.png) [@Zhengcong\_Yin](https://discuss.elastic.co/u/Zhengcong_Yin)\
**Post date:** [December 7, 2017, 6:33pm UTC](https://discuss.elastic.co/t/topic/110732/1 "2017-12-07T18:33:36Z")

</div>

43G数据， 4个主分片，没有副本。  
集群1： 单节点  
集群2： 3个主节点，4个数据节点，1个客户节点。

2种查询： sort by distance; match query

同步发送请求，计算平均的took值。 发现集群2 took值大于集群1.

如何测试水平扩展性？用什么指标及工具？ 谢谢

---

<div class="post-metadata">

**Author:** ![medcl.net](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/medcl.net/32/4414_2.png) [@medcl.net](https://discuss.elastic.co/u/medcl.net)\
**Post date:** [December 19, 2017, 2:05am UTC](https://discuss.elastic.co/t/topic/110732/2 "2017-12-19T02:05:31Z")

</div>

可以试试 Rally，这里有篇文章介绍：  
[https://elasticsearch.cn/article/275](https://elasticsearch.cn/article/275)

---

<div class="post-metadata">

**Author:** ![Zhengcong\_Yin](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/zhengcong_yin/32/62960_2.png) [@Zhengcong\_Yin](https://discuss.elastic.co/u/Zhengcong_Yin)\
**Post date:** [December 20, 2017, 6:14pm UTC](https://discuss.elastic.co/t/topic/110732/3 "2017-12-20T18:14:52Z")

</div>

感谢你的回复。

我看了一些别人的文档及书籍。都提到了增加副本可以增加load performance. 不知您是否对此有过测试？

我用Jmeter测试了一下，我每个数据节点都有完整的一套数据，并没有看到提高了load performance。不知道是不是没有路由的原因

---

<div class="post-metadata">

**Author:** ![medcl.net](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/medcl.net/32/4414_2.png) [@medcl.net](https://discuss.elastic.co/u/medcl.net)\
**Post date:** [January 15, 2018, 2:50am UTC](https://discuss.elastic.co/t/topic/110732/4 "2018-01-15T02:50:18Z")

</div>

> [@Zhengcong\_Yin](#):
>
> performance

增加副本可以提升吞吐，从而实现水平扩容，查询性能不会有提升。

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\
**Post date:** [February 12, 2018, 2:50am UTC](https://discuss.elastic.co/t/topic/110732/5 "2018-02-12T02:50:23Z")

</div>

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.