如果清单仅是本地主机，则与多个组共享的ansible变量 [英] ansible variables shared with multiple groups if inventory is only localhost

问题描述

我使用ansible通过本地计算机上的kubectl命令将作业/配置发送到我的k8s集群.我已经设置了清单文件，以便每个群集都是它自己的组，并且每个群集基本上都是到localhost的连接.

I use ansible to send jobs / configurations to my k8s cluster via the kubectl command on my local machine. I have my inventory file setup so that each cluster is it's own group and each cluster is basically a connection to localhost .

# Inventory File
#
[east.k.example.com]
localhost              ansible_connection=local

[east2.k.example.com]
localhost              ansible_connection=local

然后在我的group_vars目录中，有一个与我的清单文件不同的文件，该文件的名称与我的清单文件相同，该文件包含每个群集的所有不同变量.

Then in my group_vars directory I have a different file with the name of my group from my inventory file that holds all the different variables for each cluster.

我使用limit选项将运行限制为仅针对一个群集: ansible-playbook -vv create.yaml -l east2.k.example.com --tags ingress-generate-only

I limit my runs to target only one cluster with the limit option: ansible-playbook -vv create.yaml -l east2.k.example.com --tags ingress-generate-only

问题是，当我尝试在模板中使用变量时，我从其他组中获取了变量.我在想，因为每个组都包括本地主机.

The problem is that when I attempt to use variables in my templates I get variables from the other groups. I'm thinking because each group includes localhost.

是否有更好的方法来解决此问题?我可以设置一个标志，使组仅包含group_var文件中的变量吗?

Is there a better way to solve this issue? Can I set a flag so that groups only include the variables from the group_var files?

谢谢

推荐答案

重构广告资源以使用不同的名称:

Refactor your inventory to use distinct names:

# Inventory File
#
[east.k.example.com]
east ansible_connection=local

[east2.k.example.com]
east2 ansible_connection=local

这样，Ansible会将它们视为不同的主机，因此不会合并来自不同组的变量.

This way Ansible will treat them as different hosts, thus not merging variables from different groups.

这篇关于如果清单仅是本地主机，则与多个组共享的ansible变量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！