用于kops部署的自动现货定价

问题描述

我已经可以使用kops进行现场部署，但是它需要手动编辑实例组(节点)

I can already do spot deployments with kops but it requires manually editing a the instance groups (nodes)

$ kops edit ig --name=test.dev.test.com nodes
  machineType: t2.medium
  maxSize: 2
  minSize: 2
=>
  machineType: t1.nano
  maxSize: 1
  minSize: 1

需要研究一种以平均现货价格+ 10％自动完成此操作的方法

Need to look into a way of doing this automatically with the average spot price + 10%

我还希望至少有1个主节点和1个在正常实例上运行的节点，以在完全现货竞价关闭后生存，其余价格保持现货价格.

I would also like to have at least 1 master and 1 node that are running on normal instances to survive a complete spot-overbid shutdown and the rest to spot price.

有人可以帮我吗?

推荐答案

您可以使用Argo minion-manager.这是Kubernetes Argo 的开源工作流引擎的一部分，但可以在任何kubernetes集群上运行. Minion-manager作为部署运行，并使用AWS定价API定期更新现货实例价格.

You could use Argo minion-manager. This is part of open-source workflow engine for Kubernetes Argo , but can be run on any kubernetes cluster. Minion-manager runs as deployment and periodically updates spot instance prices using the AWS pricing API.

有关minion-manager的更多信息，请参见: https://blog.argoproj.io/use-spot-instances-with-your-kubernetes-clusters-on-aws-2a27f1887bb0

More information about minion-manager is available here: https://blog.argoproj.io/use-spot-instances-with-your-kubernetes-clusters-on-aws-2a27f1887bb0

这篇关于用于kops部署的自动现货定价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！