一.首先了解Kubernetes核心概率和功能：

Kubernetes集群包含有節(jié)點代理kubelet和Master組件(APIs, scheduler, etc)，一切都基于分布式的存儲系統(tǒng)。下面這張圖是Kubernetes的架構(gòu)圖。

 

二.單機環(huán)境部署

1.配置yum源：

wget http://mirrors.163.com/.help/CentOS7-Base-163.repo  （另外設(shè)置yum源方式y(tǒng)um-config-manager --add-repo https://download.docker.com/linux/centos/docker-ce.repo）

 

查看服務(wù)器releasever和basearch

[root@docker yum.repos.d]# rpm -qi centos-release

Name        : centos-release

Version     : 7

Release     : 1.1503.el7.centos.2.8

Architecture: x86_64

Install Date: Thu 05 Jul 2018 02:27:56 AM EDT

Group       : System Environment/Base

Size        : 34409

License     : GPLv2

Signature   : RSA/SHA256, Tue 31 Mar 2015 06:28:11 PM EDT, Key ID 24c6a8a7f4a80eb5

Source RPM  : centos-release-7-1.1503.el7.centos.2.8.src.rpm

Build Date  : Tue 31 Mar 2015 06:27:03 PM EDT

Build Host  : worker1.bsys.centos.org

Relocations : (not relocatable)

Packager    : CentOS BuildSystem <http://bugs.centos.org>

Vendor      : CentOS

Summary     : CentOS Linux release file

Description :

CentOS Linux release files

 

查看 basearch

x86_64

 

替換releasever版本

sed -i 's/$releasever/7/g' CentOS7-Base-163.repo

sed -i 's/$basearch/x86_64/g' CentOS7-Base-163.repo

 

2.關(guān)閉centos自帶防火墻和selinux

systemctl disable firewalld

systemctl stop firewalld

 

3.yum install -y etcd kubernetes

 

4.啟動相應(yīng)服務(wù)：

systemctl start etcd

systemctl start docker

 

啟動docker 服務(wù)報錯

/usr/lib/systemd/system/docker.service; disabled; vendor preset: disabled  refer#http://www.rzrgm.cn/amoyzhu/p/5261393.html

 

 

使用如下方式排錯

curl -fsSL https://get.docker.com/ | sh

# Executing docker install script, commit: 36b78b2

Warning: the "docker" command appears to already exist on this system.

 

If you already have Docker installed, this script can cause trouble, which is

why we're displaying this warning and provide the opportunity to cancel the

installation.

 

刪除上次docker目錄：

rm -rf /var/lib/docker/*

重新啟動dokcer服務(wù)正常。

 

三.使用kubernets創(chuàng)建RC和service

1.創(chuàng)建mysql-rc.yaml（參考kubernets說明文檔編寫mysql-rc.yaml文件https://kubernetes.io/docs/concepts/workloads/controllers/replicationcontroller）

kubectl create -f mysql-rc.yaml

 

2.查看POD狀態(tài)：

kubectl get pods

出現(xiàn)報錯No resources found.

 

檢查RC狀態(tài)

kubectl describe ReplicationController/mysql

Name:           mysql

Namespace:      default

Image(s):       mysql

Selector:       app=mysql

Labels:         app=mysql

Replicas:       0 current / 3 desired

Pods Status:    0 Running / 0 Waiting / 0 Succeeded / 0 Failed

No volumes.

Events:

  FirstSeen     LastSeen        Count   From                            SubObjectPath   Type            Reason          Message

  ---------     --------        -----   ----                            -------------   --------        ------          -------

  18m           0s              60      {replication-controller }                       Warning         FailedCreate    Error creating: No API token found for service account "default", retry after the token is automatically created and added to the service account

 

 

解決辦法是編輯/etc/kubernetes/apiserver 去除 KUBE_ADMISSION_CONTROL中的SecurityContextDeny,ServiceAccount，并重啟kube-apiserver.service服務(wù)： #vim /etc/kubernetes/apiserver KUBE_ADMISSION_CONTROL="--admission_control=NamespaceLifecycle,NamespaceExists,LimitRanger,ResourceQuota" #systemctl restart kube-apiserver.service 之后重新創(chuàng)建pod:

 

3.重新創(chuàng)建POC服務(wù)后，POC創(chuàng)建成果。查看pods一直處于ContainerCreating狀態(tài)

[root@docker ~]# kubectl get pods

NAME          READY     STATUS              RESTARTS   AGE

mysql-5pffq   0/1       ContainerCreating   0          5h

mysql-93zg9   0/1       ContainerCreating   0          5h

mysql-h7frf   0/1       ContainerCreating   0          5h

 

docker未運行mysql進程

[root@docker ~]# docker ps |grep mysql

 

一直處于ContainerCreating狀態(tài)，開始查找原因：

[root@docker ~]# kubectl describe pod mysql-5pffq

Name:           mysql-5pffq

Namespace:      default

Node:           127.0.0.1/127.0.0.1

Start Time:     Mon, 09 Jul 2018 13:02:43 -0400

Labels:         app=mysql

Status:         Pending

IP:

Controllers:    ReplicationController/mysql

Containers:

  mysql:

    Container ID:

    Image:                      mysql

    Image ID:

    Port:                       3306/TCP

    State:                      Waiting

      Reason:                   ContainerCreating

    Ready:                      False

    Restart Count:              0

    Volume Mounts:              <none>

    Environment Variables:      <none>

Conditions:

  Type          Status

  Initialized   True

  Ready         False

  PodScheduled  True

No volumes.

QoS Class:      BestEffort

Tolerations:    <none>

Events:

  FirstSeen     LastSeen        Count   From                    SubObjectPath   Type            Reason          Message

  ---------     --------        -----   ----                    -------------   --------        ------          -------

  5h            1m              68      {kubelet 127.0.0.1}                     Warning         FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with ErrImagePull: "image pull failed for registry.access.redhat.com/rhel7/pod-infrastructure:latest, this may be because there are no credentials on this request.  details: (open /etc/docker/certs.d/registry.access.redhat.com/redhat-ca.crt: no such file or directory)"

 

  5h    12s     1420    {kubelet 127.0.0.1}             Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with ImagePullBackOff: "Back-off pulling image \"registry.access.redhat.com/rhel7/pod-infrastructure:latest\""

 

 

嘗試手動pull image：

[root@docker ~]# docker pull registry.access.redhat.com/rhel7/pod-infrastructure:latest

Trying to pull repository registry.access.redhat.com/rhel7/pod-infrastructure ...

open /etc/docker/certs.d/registry.access.redhat.com/redhat-ca.crt: no such file or directory

 

進入：

/etc/docker/certs.d/registry.access.redhat.com/

 

查看到有個文件redhat-ca.crt -> /etc/rhsm/ca/redhat-uep.pem 一直在閃

 

[root@docker registry.access.redhat.com]# yum whatprovides */redhat-uep.pem

Loaded plugins: fastestmirror

Loading mirror speeds from cached hostfile

python-rhsm-certificates-1.19.10-1.el7_4.x86_64 : Certificates required to communicate with a Red Hat Unified Entitlement

                                                : Platform

Repo        : base

Matched from:

Filename    : /etc/rhsm/ca/redhat-uep.pem

 

安裝rhsm

yum install -y python-rhsm-certificates-1.19.10-1.el7_4.x86_64

 

重新查看POD狀態(tài)：

[root@docker registry.access.redhat.com]# kubectl get pod

NAME          READY     STATUS              RESTARTS   AGE

mysql-5pffq   0/1       ContainerCreating   0          5h

mysql-93zg9   0/1       ContainerCreating   0          5h

mysql-h7frf   0/1       ContainerCreating   0          5h

 

還是證書不存在：

[root@docker registry.access.redhat.com]# docker pull registry.access.redhat.com/rhel7/pod-infrastructure:latest

Trying to pull repository registry.access.redhat.com/rhel7/pod-infrastructure ...

open /etc/docker/certs.d/registry.access.redhat.com/redhat-ca.crt: no such file or directory

 

安裝完rhsm后證書依舊不存在

查看需要安裝的包python-rhsm-certificates-1.19.10-1.el7_4.x86_64被subscription-manager-rhsm-certificates-1.20.11-1.el7.centos.x86_64替代沒有生成證書文件

 

從centos鏡像網(wǎng)站下載對應(yīng)文件：

wget  http://mirror.centos.org/centos/7/os/x86_64/Packages/python-rhsm-certificates-1.19.10-1.el7_4.x86_64.rpm

 

手動安裝：

rpm -ivh python-rhsm-certificates-1.19.10-1.el7_4.x86_64

 

再次查看/etc/rhsm/ca/redhat-uep.pem文件已經(jīng)生成

ll /etc/rhsm/ca/redhat-uep.pem

-rw-r--r-- 1 root root 7732 Oct 19  2017 /etc/rhsm/ca/redhat-uep.pem

 

[root@docker 7]# docker pull registry.access.redhat.com/rhel7/pod-infrastructure:latest

Trying to pull repository registry.access.redhat.com/rhel7/pod-infrastructure ...

latest: Pulling from registry.access.redhat.com/rhel7/pod-infrastructure

26e5ed6899db: Already exists

66dbe984a319: Already exists

9138e7863e08: Already exists

Digest: sha256:92d43c37297da3ab187fc2b9e9ebfb243c1110d446c783ae1b989088495db931

Status: Image is up to date for registry.access.redhat.com/rhel7/pod-infrastructure:latest

 

4.創(chuàng)建mysql service：

kubectl  create -f mysql-svc.yaml

查看POD狀態(tài)處于Pending 狀態(tài)：

[root@docker ~]# kubectl describe pod mysql-1cg54

Name:           mysql-1cg54

Namespace:      default

Node:           127.0.0.1/127.0.0.1

Start Time:     Mon, 09 Jul 2018 19:24:31 -0400

Labels:         app=mysql

Status:         Pending

IP:

Controllers:    ReplicationController/mysql

Containers:

  mysql:

    Container ID:

    Image:                      mysql

    Image ID:

    Port:                       3306/TCP

    State:                      Waiting

      Reason:                   ContainerCreating

    Ready:                      False

    Restart Count:              0

    Volume Mounts:              <none>

    Environment Variables:      <none>

Conditions:

  Type          Status

  Initialized   True

  Ready         False

  PodScheduled  True

No volumes.

QoS Class:      BestEffort

Tolerations:    <none>

Events:

  FirstSeen     LastSeen        Count   From                    SubObjectPath   Type            Reason          Message

  ---------     --------        -----   ----                    -------------   --------        ------          -------

  2m            2m              1       {default-scheduler }                    Normal          Scheduled       Successfully assigned mysql-1cg54 to 127.0.0.1

  2m            2m              1       {kubelet 127.0.0.1}                     Warning         FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with RunContainerError: "runContainer: Error response from daemon: {\"message\":\"error creating overlay mount to /var/lib/docker/overlay2/e59f2093257452779d9a17bcd555409d7e6b0492e2530d426494003083b081ff-init/merged: invalid argument\"}"

 

  1m    1m      1       {kubelet 127.0.0.1}             Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with RunContainerError: "runContainer: Error response from daemon: {\"message\":\"error creating overlay mount to /var/lib/docker/overlay2/838d7c94495bcfbad3b668966a44acbd7e6468b3356835e47dc3d6325aa9b948-init/merged: invalid argument\"}"

 

  1m    1m      1       {kubelet 127.0.0.1}             Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with RunContainerError: "runContainer: Error response from daemon: {\"message\":\"error creating overlay mount to /var/lib/docker/overlay2/9f00835eaa13507b4bf2674dec732d8814b076a989db644510c81e8442564e5e-init/merged: invalid argument\"}"

 

  1m    1m      1       {kubelet 127.0.0.1}             Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with RunContainerError: "runContainer: Error response from daemon: {\"message\":\"error creating overlay mount to /var/lib/docker/overlay2/d7ff3025c0a133856be50007c53dde4e6ec87daa2d62a97cc6ce3494a3e641f4-init/merged: invalid argument\"}"

 

  1m    1m      1       {kubelet 127.0.0.1}             Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with RunContainerError: "runContainer: Error response from daemon: {\"message\":\"error creating overlay mount to /var/lib/docker/overlay2/eb14878a8937c5d87cd83b19c51dd85ab5c7bff5cf93d367e8a55d36ef8c18a6-init/merged: invalid argument\"}"

 

  52s   52s     1       {kubelet 127.0.0.1}             Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with RunContainerError: "runContainer: Error response from daemon: {\"message\":\"error creating overlay mount to /var/lib/docker/overlay2/cf31456fd3c3bddba7997c030ca4a9ead718711c2cd447005cd4214abe619788-init/merged: invalid argument\"}"

 

  39s   39s     1       {kubelet 127.0.0.1}             Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with RunContainerError: "runContainer: Error response from daemon: {\"message\":\"error creating overlay mount to /var/lib/docker/overlay2/9cfe8f28521d55f020fbb5e02c10a6604d82800c4e48d1aac4c66ec71fd4edf2-init/merged: invalid argument\"}"

 

  28s   28s     1       {kubelet 127.0.0.1}             Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with RunContainerError: "runContainer: Error response from daemon: {\"message\":\"error creating overlay mount to /var/lib/docker/overlay2/c044367cb344a42a9df4fe809344950e575013b9bcc920416a53c489bd5f74dd-init/merged: invalid argument\"}"

 

  17s   17s     1       {kubelet 127.0.0.1}             Warning FailedSync      Error syncing pod, skipping: failed to "StartContainer" for "POD" with RunContainerError: "runContainer: Error response from daemon: {\"message\":\"error creating overlay mount to /var/lib/docker/overlay2/e852c1fef7eed8db0567052818a4f4f71bdffccc216a747d92ea0d78492347f1-init/merged: invalid argument\"}"

 

  2m    5s      10      {kubelet 127.0.0.1}             Warning MissingClusterDNS       kubelet does not have ClusterDNS IP configured and cannot create Pod using "ClusterFirst" policy. Falling back to DNSDefault policy.

  5s    5s      1       {kubelet 127.0.0.1}             Warning FailedSync              (events with common reason combined)

 

 

報錯是因為用的overlay2文件系統(tǒng)，而系統(tǒng)默認只能識別overlay文件系統(tǒng)

所以我們就要更新文件系統(tǒng)了

例：DOCKER_STORAGE_OPTIONS="--storage-driver overlay "

 

再次創(chuàng)建mysql service：

kubectl  create -f mysql-svc.yaml