Feature: when master0 breaks down, other masters can use sealos cluster management commands #3859
Comments
|
/assign |
/assign |
|
用kubeadm取代sealos。 这个方式不可行，只需要同步配置即可。 |
Replace sealos with kubeadm. This method is not feasible, only the configuration needs to be synchronized. |
添加switch命令，取代异常的master0, 例如：sealos switch --master 192.168.0.5 。 |
Add the switch command to replace the abnormal master0, for example: sealos switch --master 192.168.0.5. |
|
不需要新添加命令，在集群初始化和加入节点操作成功后同步master ~/.sealos数据就可以的 |
No need to add new commands, just synchronize the master ~/.sealos data after the cluster initialization and node joining operations are successful |
我明白了，保持master节点之间的~/.sealos数据同步。 |
I get it, keep ~/.sealos data in sync between master nodes. |
|
我们也有同样需求，我觉得仓库信息/var/lib/sealos/data/default/rootfs/registry/ 等数据也需要同步的 |
We also have the same requirement, I think warehouse information /var/lib/sealos/data/default/rootfs/registry/ and other data also need to be synchronized |
这个当前支持的，你安装集群的时候 设置role=registry 就可以了 |
This is currently supported, you can set role=registry when you install the cluster |
|
能否详细介绍一下如何设置role=registry，或者提供文档？谢谢 |
Can you explain in detail how to set role=registry, or provide documentation? Thanks |
|
sealos gen 然后把节点role添加 registry 然后apply即可。 |
sealos gen and then add the node role to registry and then apply. |
|
apiVersion: apps.sealos.io/v1beta1 |
是的，每次安装app会往三个仓库发送镜像。可能会慢，但是如果master0坏了 你只要把每个节点的sealos.hub对应的IP改掉就好了 |
Yes, every time the app is installed, images will be sent to three warehouses. It may be slow, but if master0 is broken, you only need to change the IP corresponding to the sealos.hub of each node. |
|
好的，这个功能是从sealos哪个版本开始支持的？ |
Ok, from which version of sealos is this function supported? |
#2096 这个PR |
#2096 this PR |
|
好的，谢谢。
这个功能是需要新开发的吧？ |
OK, thanks.
Does this function need to be newly developed? |
是的，也可以手动执行 sealos scp -r master src target 命令来同步~/.sealos数据到其他master节点 |
Yes, you can also manually execute the sealos scp -r master src target command to synchronize ~/.sealos data to other master nodes |
|
This issue has been automatically closed because we haven't heard back for more than 60 days, please reopen this issue if necessary. |
|
请问这个功能做完了吗？是否计划合入到新的版本中？谢谢 |
Is this function completed? Are there plans to incorporate it into the new version? Thanks |
|
最新版应该已经支持了 @gogogo2156 |
The latest version should already support @gogogo2156 |
|
@bxy4543 最新版是指v4.3.7吗？需要去配置开启吗？谢谢 |
Does the latest version refer to v4.3.7? Do I need to configure it to enable it? Thanks |
|
怎么在安装时，用命令指定部署3个registry |
How to use the command to specify the deployment of three registries during installation |
What is the problem this feature will solve?
当master0发生故障时，其他master可以使用sealos集群管理命令。
When master0 breaks down, other masters can use sealos cluster management commands.
我的master0发生故障，无法恢复。但是其他master不能使用sealos集群管理命令。即使在其他master安装sealos工具（apt install sealos）也因缺失~/.sealos/和/var/lib/sealos/data/default/rootfs/registry/等重要信息而无法正常工作。
My master0 breaks down and cannot be recovered. But other masters cannot use sealos cluster management commands. Even if the sealos tool is installed on other masters (apt install sealos), it cannot function properly due to missing important information such as ~/. sealos/ and /var/lib/sealos/data/default/rootfs/registry/.
If you have solution，please describe it
目前没有解决方案，只能重建集群。
No solution for now but rebuild cluster.
What alternatives have you considered?
用kubeadm取代sealos。
Replace sealos with kubeadm.
The text was updated successfully, but these errors were encountered: