Kubeadm

Question

尝试使用 kubeadm 在 3 个 Debian 10 VM 上配置 k8s 集群。

所有虚拟机都有 2 个网络接口，eth0 作为 public 静态 ip 接口，eth1 作为本地接口，静态 ips 在 192.168.0.0/16:

大师：192.168.1.1
节点 1：192.168.2.1
节点 2：192.168.2.2

所有节点之间都有互连。

ip a 来自主控主机：

# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 00:52:70:53:d5:12 brd ff:ff:ff:ff:ff:ff
    inet XXX.XXX.244.240/24 brd XXX.XXX.244.255 scope global dynamic eth0
       valid_lft 257951sec preferred_lft 257951sec
    inet6 2a01:367:c1f2::112/48 scope global 
       valid_lft forever preferred_lft forever
    inet6 fe80::252:70ff:fe53:d512/64 scope link 
       valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 00:95:af:b0:8c:c4 brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.1/16 brd 192.168.255.255 scope global eth1
       valid_lft forever preferred_lft forever
    inet6 fe80::295:afff:feb0:8cc4/64 scope link 
       valid_lft forever preferred_lft forever

主节点初始化良好：

kubeadm init --upload-certs --apiserver-advertise-address=192.168.1.1 --apiserver-cert-extra-sans=192.168.1.1,XXX.XXX.244.240 --pod-network-cidr=10.40.0.0/16 -v=5

Output

但是当我加入工作节点时，kube-api 无法访问：

kubeadm join 192.168.1.1:6443 --token 7bl0in.s6o5kyqg27utklcl --discovery-token-ca-cert-hash sha256:7829b6c7580c0c0f66aa378c9f7e12433eb2d3b67858dd3900f7174ec99cda0e -v=5

Output

来自主人的 Netstat：

# netstat -tupn | grep :6443
tcp        0      0 192.168.1.1:43332       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:41774       192.168.1.1:6443        ESTABLISHED 5362/kube-proxy     
tcp        0      0 192.168.1.1:41744       192.168.1.1:6443        ESTABLISHED 5236/kubelet        
tcp        0      0 192.168.1.1:43376       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:43398       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:41652       192.168.1.1:6443        ESTABLISHED 4914/kube-scheduler 
tcp        0      0 192.168.1.1:43448       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:43328       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:43452       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:43386       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:43350       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:41758       192.168.1.1:6443        ESTABLISHED 5182/kube-controlle 
tcp        0      0 192.168.1.1:43306       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:43354       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:43296       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:43408       192.168.1.1:6443        TIME_WAIT   -                   
tcp        0      0 192.168.1.1:41730       192.168.1.1:6443        ESTABLISHED 5182/kube-controlle 
tcp        0      0 192.168.1.1:41738       192.168.1.1:6443        ESTABLISHED 4914/kube-scheduler 
tcp        0      0 192.168.1.1:43444       192.168.1.1:6443        TIME_WAIT   -                   
tcp6       0      0 192.168.1.1:6443        192.168.1.1:41730       ESTABLISHED 5094/kube-apiserver 
tcp6       0      0 192.168.1.1:6443        192.168.1.1:41744       ESTABLISHED 5094/kube-apiserver 
tcp6       0      0 192.168.1.1:6443        192.168.1.1:41738       ESTABLISHED 5094/kube-apiserver 
tcp6       0      0 192.168.1.1:6443        192.168.1.1:41652       ESTABLISHED 5094/kube-apiserver 
tcp6       0      0 ::1:6443                ::1:42862               ESTABLISHED 5094/kube-apiserver 
tcp6       0      0 192.168.1.1:6443        192.168.1.1:41758       ESTABLISHED 5094/kube-apiserver 
tcp6       0      0 ::1:42862               ::1:6443                ESTABLISHED 5094/kube-apiserver 
tcp6       0      0 192.168.1.1:6443        192.168.1.1:41774       ESTABLISHED 5094/kube-apiserver

Pods 来自大师：

# kubectl --kubeconfig=/etc/kubernetes/admin.conf get pods -n kube-system -o wide
NAME                                              READY   STATUS    RESTARTS   AGE   IP                   NODE                      NOMINATED NODE   READINESS GATES
coredns-558bd4d5db-8qhhl                          0/1     Pending   0          12m   <none>               <none>                    <none>           <none>
coredns-558bd4d5db-9hj7z                          0/1     Pending   0          12m   <none>               <none>                    <none>           <none>
etcd-cloud604486.fastpipe.io                      1/1     Running   0          12m   2a01:367:c1f2::112   cloud604486.fastpipe.io   <none>           <none>
kube-apiserver-cloud604486.fastpipe.io            1/1     Running   0          12m   2a01:367:c1f2::112   cloud604486.fastpipe.io   <none>           <none>
kube-controller-manager-cloud604486.fastpipe.io   1/1     Running   0          12m   2a01:367:c1f2::112   cloud604486.fastpipe.io   <none>           <none>
kube-proxy-dzd42                                  1/1     Running   0          12m   2a01:367:c1f2::112   cloud604486.fastpipe.io   <none>           <none>
kube-scheduler-cloud604486.fastpipe.io            1/1     Running   0          12m   2a01:367:c1f2::112   cloud604486.fastpipe.io   <none>           <none>

所有虚拟机都设置了这个内核参数：

{ name: 'vm.swappiness', value: '0' }
{ name: 'net.bridge.bridge-nf-call-iptables', value: '1' }
{ name: 'net.bridge.bridge-nf-call-ip6tables', value: '1'}
{ name: 'net.ipv4.ip_forward', value: 1 }
{ name: 'net.ipv6.conf.all.forwarding', value: 1}

br_netfilter 内核模块处于活动状态并且 iptables 设置为旧模式（通过替代方案）

我是不是漏掉了什么？

Answer 1

您遇到问题的原因是必须确保组件之间的 TLS 连接安全。从 kubelet 的角度来看，如果 Api-server 证书将在替代名称中包含我们要连接的服务器的 IP，这将是安全的。您可以自己注意到，您只添加到 SANs 一个 IP 地址。

如何解决这个问题？有两种方式：

在您的节点中将 --discovery-token-unsafe-skip-ca-verification 标志与您的 kubeadm 加入命令一起使用。
在集群初始化阶段（kubeadm init）将第二个NIC的IP地址添加到SANsapi证书

如需更多阅读，请查看此直接相关的 PR #93264，它在 kubernetes 1.19 中引入。

Answer 2

经过 1 周的修补，问题归结为服务提供商网络配置错误。

对于遇到同样问题的任何人，请检查您网络的 MTU，在我的情况下，它默认为 1500 而不是推荐的 1450。

Kubeadm - 无法加入节点 - 请求在等待连接时被取消

Kubeadm - unable to join nodes - request canceled while waiting for connection

kubernetes