dyrnq

kubeadm-certs has an expiration binding token(default 1 hour), when after 1 hour, the kubeadm-certs will auto gone, there has two method solve: 1、re upload kubeadm-certs agagin on first master by`kubeadm...

Look like your IP 10.0.20.30 does not have route.You must create link device by manual on every node first. ```bash ip link add lvscare type dummy ip addr add 10.0.20.30/32...

You can refer to kube-scheduler POD && kube-controller POD installed and configed by kubeadm they have host `localhost` and `127.0.0.1` at least check on my k8s cluster installed by kubeadm...

**supplement** there have two kind of cert : one is client of kube-apiserver (configed in /etc/kubernetes/controller-manager.conf client-certificate-data:) antoher is self server tls by options `--tls-cert-file` and `--tls-private-key-file` sometimes they can...

YES，`self server tls` SAN of localhost and 127.0.0.1 for **https health check**

文档里写的是写到 `--config=/etc/kubernetes/kubelet-config.yaml`的

cloud plugin support tencent polarismesh

Recommend use [wait4x](https://github.com/atkrad/wait4x) or [wait-for-it](https://github.com/vishnubob/wait-for-it). > 在`docker-production.online.yml`中我启用两个`php`服务`php7`和`php8`，并启用一个`nginx`服务，将`nginx`服务依赖于`php`： > > ``` > depends_on: > - php7 > - php8 > ``` > > 在启动服务的时候`$ ./lnmp-docker swarm-deploy`，有的时候并非按照顺序执行，就会造成启动服务失败 > > ``` > Creating...

Client side load balancing with fault tolerance, support multiple protocol (HTTP, TCP, UDP). like Netflix/ribbon and Spring Cloud LoadBalancer.

> Reference in new 谢谢，我看看 我发现一些plugin 比如 https://gitee.com/noear/solon/tree/master/_solon_plugin_cloud/zookeeper-solon-cloud-plugin 这里的 ```bash solon.cloud.zookeeper: server: localhost #服务地址 ``` 的这里连接 zookeeper的地址是否可以传入多个server 地址然后实现client side的LB 还有 https://gitee.com/noear/solon/tree/master/_solon_plugin_cloud/nacos2-solon-cloud-plugin 这里的 ```bash solon.cloud.nacos2: server: localhost:8848 #服务地址 ``` 的server，一般这种基础中间件都是集群部署，他们或是tcp或是grcp等协议，那么就需要一个VIP来负载均衡nacos他们，假如这里可以填写多个地址的话，相当于在client side...