Cassandra负载平衡与有序分区器？ [英] Cassandra load balancing with an ordered partitioner?

问题描述

所以我在这里看到Cassandra没有自动负载平衡，这在使用有序分区器时会出现（一组行的某些常见的值范围将被存储在相对较少的机器上，的查询）。

在设计Cassandra数据模型时最好的做法是什么？

我还是Cassandra的新手，它是如何工作的。如何避免这个问题，所以范围查询仍然可能？

解决方案

正如我们在上面提到的其他帖子，Cassandra 0.5支持半自动负载平衡，您所要做的是告诉节点负载平衡，它会自动移动到令牌环上的繁忙地方。

这在 http://wiki.apache.org/cassandra/Operations 中进行了说明

So I see here that Cassandra does not have automatic load balancing, which comes into view when using the ordered partitioner (a certain common range of values of a group of rows would be stored on a relatively few machines which would then serve most of the queries).
What's The Best Practice In Designing A Cassandra Data Model?

I'm still new to Cassandra and how it works. how would one go about avoiding this issue, so that range queries are still possible? I didn't really get the above answers (linked url) idea about appending a hash to keys.

解决方案

As mentioned on the other post, Cassandra 0.5 supports semiautomatic load balancing, where all you have to do is tell a node to loadbalance and it will move to a busier place on the token ring automatically.

This is covered in http://wiki.apache.org/cassandra/Operations

这篇关于Cassandra负载平衡与有序分区器？的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！