本地多区域Kubernetes联合集群 [英] On-premise multi-zone Kubernetes federation cluster

查看:77
本文介绍了本地多区域Kubernetes联合集群的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

假设我在纽约,伦敦,德里,北京拥有 baremetal服务器.我的要求是在分布式环境中加入所有4台裸机服务器并运行服务在kubernetes上. 我该如何实现?

Let's say I have baremetal servers at New York, London, Del Beijing. My requirement is to join all the 4 baremetal servers in a distributed environment and run services on top of kubernetes. How do I achieve this ?

推荐答案

问题过于广泛,以下是我对全球数据存储的见解:

Question is too broad but here's my take with insights on a global datastore:

选项1:为每个区域设置一个K8s集群,并使它们通过服务相互通信.让每个人都与自己的数据存储区进行对话,并使用

Option 1: Setup a K8s cluster for each region and have them talk to each other through services. Have each talk to their own datastores, and keep data separate per region an use something like GSLB to route your traffic.

选项2.为每个区域设置一个K8s集群,并使用一个全局数据库,例如 Dynamdb Global表格 Cloud Spanner

Option 2. Setup a K8s cluster for each region and use a global database like Dynamdb Global Tables, Cloud Spanner or CosmosDB.

选项3.使用 Kubernetes联盟.联合不一定能解决多区域数据存储的难题,在撰写本文时,它也处于Beta版.但是,它将有助于跨多个区域管理您的Kubernetes服务.

Option 3. Use Kubernetes Federation. Federation doesn't necessarily solve the multi-region data store challenge, and it's also in Beta as of this writing. However, it will help manage your Kubernetes services across multiple regions.

选项4.您自己创建的其他东西很有趣!好玩!

Option 4. Something else, that you create on your own, fun! fun!

这篇关于本地多区域Kubernetes联合集群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!

查看全文
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆