docker&k8s填坑记

本篇主要用于记录在实施docker和kubenetes过程中遇到的一个问题和解决办法.

kubernets nodeport 无法访问

环境:

Os: centos7.1
Kubelet: 1.6.7
Docker: 17.06-ce
Calico: 2.3

K8s Cluster: master, node-1, node-2

问题:

现有 service A， 为了能使外部访问，故将 service type 设为NodePort。端口为 31246。
A 所对应的 pod 运行在 node-1 上。

经过测试发现，外部访问 master:31246 和 node-2:31246 时均出现失败，只有通过 node-1:31246 才可正常访问。

起因:

为了安全起见， docker 在 1.13 版本之后，将系统iptables 中 FORWARD 链的默认策略设置为 DROP，并为连接到 docker0 网桥的容器添加了放行规则。这里引用 moby issue#14041 中的描述：

When docker starts, it enables net.ipv4.ip_forward without changing the iptables FORWARD chain default policy to DROP. This means that another machine on the same network as the docker host can add a route to their routing table, and directly address any containers running on that docker host.

For example, if the docker0 subnet is 172.17.0.0/16 (the default subnet), and the docker host’s IP address is 192.168.0.10, from another host on the network run:

$ ip route add 172.17.0.0/16 via 192.168.0.10
$ nmap 172.17.0.0/16

1
2

The above will scan for containers running on the host, and report IP addresses & running services found.

To fix this, docker needs to set the FORWARD policy to DROP when it enables the net.ipv4.ip_forward sysctl parameter.

kubernetes 使用的 cni 插件会因此受影响（cni并不会在 FORWARD 链中生成相应规则），由此导致除 pod 所在 host 以外节点无法转发报文而访问失败。

解决办法:

如果对安全要求较低，可将 FORWARD 链的默认规则设为 ACCEPT

iptables -P FORWARD ACCEPT

google 网络不可达

https://packages.cloud.google.com/yum/repos/kubernetes-el7-x86_64/repodata/repomd.xml: [Errno 14] curl#7 - "Failed to connect to 2404:6800:4005:809::200e: 网络不可达"

经典问题，需要自备梯子哦！
CentOS中设置yum的proxy

vi /etc/yum.conf
# 增加内容如下：
proxy=http://xxx.xx.x.xx:xxx #代理地址

没有梯子的换阿里源

cat > /etc/yum.repos.d/kubernetes.repo <<EOF
[kubernetes]
name=Kubernetes
baseurl=https://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64
enabled=1
gpgcheck=0
repo_gpgcheck=0
EOF

关闭Swap

F1213 10:20:53.304755    2266 server.go:261] failed to run Kubelet: Running with swap on is not supported, please disable swap! or set --fail-swap-on flag to false. /proc/swaps contained:

执行

swapoff -a

设定master错误

[ERROR FileContent--proc-sys-net-bridge-bridge-nf-call-iptables]: /proc/sys/net/bridge/bridge-nf-call-iptables contents are not set to 1
[ERROR FileContent--proc-sys-net-ipv4-ip_forward]: /proc/sys/net/ipv4/ip_forward contents are not set to 1
[preflight] If you know what you are doing, you can make a check non-fatal with `--ignore-preflight-errors=...`

按提示设定为1

echo "1" >/proc/sys/net/ipv4/ip_forward
echo "1" >/proc/sys/net/bridge/bridge-nf-call-iptables

Kubeadm init 安装镜像卡住

# kubeadm config images list --kubernetes-version v1.13.0 # 看下该版本下的镜像名
# 拉取镜像
docker pull mirrorgooglecontainers/kube-apiserver:v1.13.0
docker pull mirrorgooglecontainers/kube-controller-manager:v1.13.0
docker pull mirrorgooglecontainers/kube-scheduler:v1.13.0
docker pull mirrorgooglecontainers/kube-proxy:v1.13.0
docker pull mirrorgooglecontainers/pause:3.1
docker pull mirrorgooglecontainers/etcd:3.2.24
docker pull coredns/coredns:1.2.6

# 重命名镜像标签
docker tag docker.io/mirrorgooglecontainers/kube-proxy:v1.13.0 k8s.gcr.io/kube-proxy:v1.13.0
docker tag docker.io/mirrorgooglecontainers/kube-scheduler:v1.13.0 k8s.gcr.io/kube-scheduler:v1.13.0
docker tag docker.io/mirrorgooglecontainers/kube-apiserver:v1.13.0 k8s.gcr.io/kube-apiserver:v1.13.0
docker tag docker.io/mirrorgooglecontainers/kube-controller-manager:v1.13.0 k8s.gcr.io/kube-controller-manager:v1.13.0
docker tag docker.io/mirrorgooglecontainers/etcd:3.2.24  k8s.gcr.io/etcd:3.2.24
docker tag docker.io/mirrorgooglecontainers/pause:3.1  k8s.gcr.io/pause:3.1
docker tag docker.io/coredns/coredns:1.2.6  k8s.gcr.io/coredns:1.2.6

# 删除旧镜像
docker rmi docker.io/mirrorgooglecontainers/kube-proxy:v1.13.0
docker rmi docker.io/mirrorgooglecontainers/kube-scheduler:v1.13.0
docker rmi docker.io/mirrorgooglecontainers/kube-apiserver:v1.13.0
docker rmi docker.io/mirrorgooglecontainers/kube-controller-manager:v1.13.0
docker rmi docker.io/mirrorgooglecontainers/etcd:3.2.24
docker rmi docker.io/mirrorgooglecontainers/pause:3.1
docker rmi docker.io/coredns/coredns:1.2.6

终于看到了

Your Kubernetes master has initialized successfully!

To start using your cluster, you need to run the following as a regular user:

mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config

You should now deploy a pod network to the cluster.
Run "kubectl apply -f [podnetwork].yaml" with one of the options listed at:
https://kubernetes.io/docs/concepts/cluster-administration/addons/

You can now join any number of machines by running the following on each node
as root:

kubeadm join 192.168.232.204:6443 --token m2hxkd.scxjrxgew6pyhvmb --discovery-token-ca-cert-hash sha256:8b94cefbe54ae4b3d7201012db30966c53870aad55be80a2888ec0da178c3610

network配置

# 这是我的虚机/etc/hosts的配置
192.168.232.204   k8a204
192.168.232.203 k8a203
192.168.232.202 k8a202

按手册安装选用的网络，并等待dns安装OK，然后增加node。

NAME     STATUS   ROLES    AGE    VERSION
k8a204   Ready    master   6m6s   v1.13.0
[root@localhost .kube]# kubectl get nodes
NAME     STATUS     ROLES    AGE     VERSION
k8a203   NotReady   <none>   4s      v1.13.0
k8a204   Ready      master   6m19s   v1.13.0

注意，配置较慢，耐心等待

kubectl get pods --all-namespaces
===============以下是结果===============
NAMESPACE     NAME                             READY   STATUS              RESTARTS   AGE
kube-system   coredns-86c58d9df4-2vdvx         1/1     Running             0          7m32s
kube-system   coredns-86c58d9df4-88fjk         1/1     Running             0          7m32s
kube-system   etcd-k8a204                      1/1     Running             0          6m39s
kube-system   kube-apiserver-k8a204            1/1     Running             0          6m30s
kube-system   kube-controller-manager-k8a204   1/1     Running             0          6m30s
kube-system   kube-proxy-tl7g5                 1/1     Running             0          7m32s
kube-system   kube-proxy-w2jgl                 0/1     ContainerCreating   0          95s
kube-system   kube-scheduler-k8a204            1/1     Running             0          6m49s

节点加入后NotReady

接上一问题：
ContainerCreating 状态时，请耐心等待，但是如果超过10分钟仍然无响应，则必定是出错了，囧！
最主要的问题：节点的镜像拉不下来。
采用下列方式：

1)在master主机内保存镜像为文件：

docker save -o /opt/kube-pause.tar k8s.gcr.io/pause:3.1
docker save -o /opt/kube-proxy.tar k8s.gcr.io/kube-proxy:v1.13.0
docker save -o /opt/kube-flannel1.tar quay.io/coreos/flannel:v0.9.1
docker save -o /opt/kube-flannel2.tar quay.io/coreos/flannel:v0.10.0-amd64
docker save -o /opt/kube-calico1.tar quay.io/calico/cni:v3.3.2
docker save -o /opt/kube-calico2.tar quay.io/calico/node:v3.3.2

2)拷贝文件到node计算机

scp /opt/*.tar root@192.168.232.203:/opt/

3) 在node节点执行docker导入

docker load -i /opt/kube-flannel1.tar
docker load -i /opt/kube-flannel2.tar
docker load -i /opt/kube-proxy.tar
docker load -i /opt/kube-pause.tar
docker load -i /opt/kube-calico1.tar
docker load -i /opt/kube-calico2.tar

4) 检查node节点镜像文件

docker images
==============================================以下是结果======================================
REPOSITORY               TAG                 IMAGE ID            CREATED             SIZE
k8s.gcr.io/kube-proxy    v1.13.0             8fa56d18961f        9 days ago          80.2 MB
quay.io/calico/node      v3.3.2              4e9be81e3a59        9 days ago          75.3 MB
quay.io/calico/cni       v3.3.2              490d921fa49c        9 days ago          75.4 MB
quay.io/coreos/flannel   v0.10.0-amd64       f0fad859c909        10 months ago       44.6 MB
k8s.gcr.io/pause         3.1                 da86e6ba6ca1        11 months ago       742 kB
quay.io/coreos/flannel   v0.9.1              2b736d06ca4c        13 months ago       51.3 MB

搞定了，所有服务均running

[root@localhost .kube]# kubectl get pods --all-namespaces
====================================以下是结果========================================
NAMESPACE     NAME                             READY   STATUS    RESTARTS   AGE
kube-system   calico-node-4dsg5                1/2     Running   0          42m
kube-system   calico-node-5dtk2                1/2     Running   0          41m
kube-system   calico-node-78qvp                1/2     Running   0          41m
kube-system   coredns-86c58d9df4-26vr7         1/1     Running   0          43m
kube-system   coredns-86c58d9df4-s5ljf         1/1     Running   0          43m
kube-system   etcd-k8a204                      1/1     Running   0          42m
kube-system   kube-apiserver-k8a204            1/1     Running   0          42m
kube-system   kube-controller-manager-k8a204   1/1     Running   0          42m
kube-system   kube-proxy-8c7hs                 1/1     Running   0          41m
kube-system   kube-proxy-dls8l                 1/1     Running   0          41m
kube-system   kube-proxy-t65tc                 1/1     Running   0          43m
kube-system   kube-scheduler-k8a204            1/1     Running   0          42m

重启恢复master

swapoff -a
# 启动所有容器
# 更简洁的命令： docker start $(docker ps -aq)
docker start $(docker ps -a | awk '{ print $1}' | tail -n +2)
systemctl start kubelet
# 查看启动错误
journalctl -xefu kubelet
# docker 开机自启
docker run --restart=always

• DNS解析 kubernetes.default失败

安装busybox进行dns检测，一直出现如下错误：

kubectl exec -ti  busybox -- nslookup kubernetes.default
=============================以下是结果============================
Server:         10.96.0.10
Address:        10.96.0.10:53
** server can't find kubernetes.default: NXDOMAIN
*** Can't find kubernetes.default: No answer

经查，新版busybox的dns解析有变化或bug，采用旧版本busybox images <= 1.28.4 后测试OK

• token过期后重新生成

# 生成新的token
kubeadm token create
# 生成新的token hash码
openssl x509 -pubkey -in /etc/kubernetes/pki/ca.crt | openssl rsa -pubin -outform der 2>/dev/null | openssl dgst -sha256 -hex | sed 's/^.* //'
# 利用新的token和hash码加入节点
# master地址，token，hash请自行更换
kubeadm join 192.168.232.204:6443 --token m87q91.gbcqhfx9ansvaf3o --discovery-token-ca-cert-hash sha256:fdd34ef6c801e382f3fb5b87bc9912a120bf82029893db121b9c8eae29e91c62