学习笔记
1. 安装要求
在开始之前，Dcoker Swarm 集群部署 机器需要满足以下几个条件：
- 虚拟了3台机器，操作系统：Linux version 3.10.0-1127.el7.x86_64
- 硬件配置：CPU(s)-2；3GB或更多RAM;硬盘20G左右；
- 3台机器构成一个集群，机器之前网络互通；
- 集群可以访问外网，用于拉取镜像
- 禁用swap分区
2. Swarm 的工作模式
2.1 集群调度图原理
swarm 集群由管理节点（manager）和工作节点（work node）构成。
2.2 Swarm集群调用架构图
docker 集群之间通过远程API实现通讯（Remote API）
3. 环境准备
- master节点 ： 192.168.11.99
- node1节点 ：192.168.11.131
- node2节点 ：192.168.11.130
## 1. 关闭防火墙
$ systemctl stop firewalld
### 关闭防火墙自动启动
$ systemctl disable firewalld
## 2. 关闭selinux
### 2.1 修改配置文件永久关闭selinux【重启生效】
$ vi /etc/sysconfig/selinux
SELINUX=enforcing
### 2.2 临时关闭selinux
$ setenforce 0
## 3. 分别设置设置主机名：
$ hostnamectl set-hostname swarm-master
$ hostnamectl set-hostname swarm-node1
$ hostnamectl set-hostname swarm-node2
### 3.1 查看hostname
$ hostname
## 4. 配置hosts文件
$ cat >> /etc/hosts << EOF
192.168.11.99 swarm-master
192.168.11.130 swarm-node1
192.168.11.131 swarm-node2
EOF
## 5. 安裝docker
$ yum -y install docker
### 5.1 docker服务启动
$ systemctl enable docker && systemctl start docker
4.创建Swarm并添加节点
# Swarm-master 虚拟机操作
## 1. 创建Swarm集群(记录token)
[root@localhost vagrant]# docker swarm init --advertise-addr 192.168.11.99
Swarm initialized: current node (6jir7f507ur05lrhtc2xq8plt) is now a manager.
To add a worker to this swarm, run the following command:
# 这就是添加节点的方式(要保存初始化后token，因为在节点加入时要使用token作为通讯的密钥)
docker swarm join --token SWMTKN-1-30sw7z3qrh3gs0pi5fmllypw0c5dxrnpzf2uwd5nldcifzzroh-0l4pkovtgy708qs5pn5v8rzsk 192.168.11.99:2377
To add a manager to this swarm, run 'docker swarm join-token manager' and follow the instructions.
## 2. 查看集群的相关信息
[root@localhost vagrant]# docker info | grep swarm
WARNING: You're not using the default seccomp profile
WARNING: bridge-nf-call-ip6tables is disabled
Name: swarm-master
### 2.1 查看节点信息
[root@localhost vagrant]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS
6jir7f507ur05lrhtc2xq8plt * swarm-master Ready Active Leader
# Swarm-node 虚拟机
## 3. 加入的集群
### 3.1 Swarm-node1 节点加入到集群
[root@localhost vagrant]# docker swarm join --token SWMTKN-1-30sw7z3qrh3gs0pi5fmllypw0c5dxrnpzf2uwd5nldcifzzroh-
0l4pkovtgy708qs5pn5v8rzsk 192.168.11.99:2377
This node joined a swarm as a worker.
### 3.2 Swarm-node2 节点加入到集群
[root@localhost vagrant]# docker swarm join --token SWMTKN-1-30sw7z3qrh3gs0pi5fmllypw0c5dxrnpzf2uwd5nldcifzzroh-
0l4pkovtgy708qs5pn5v8rzsk 192.168.11.99:2377
This node joined a swarm as a worker.
## 4.查看集群节点状态
[root@localhost vagrant]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS
6jir7f507ur05lrhtc2xq8plt * swarm-master Ready Active Leader
qt9ja2qbmq4n040y60g6xbdlz swarm-node2 Ready Active
scnyaecep2d1eiccvubusykww swarm-node1 Ready Active
5. Swarm 常用操作
更改节点的availablity状态
swarm集群中node的availability状态可以为 active或者drain，其中：
active状态下，node可以接受来自manager节点的任务分派；
drain状态下，node节点会结束task，且不再接受来自manager节点的任务分派（也就是下线节点）
## 1.删除节点
[root@localhost vagrant]# docker node rm --force swarm-node1
swarm-node1
[root@localhost vagrant]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS
6jir7f507ur05lrhtc2xq8plt * swarm-master Ready Active Leader
qt9ja2qbmq4n040y60g6xbdlz swarm-node2 Ready Active
## 2. 更新状态
[root@localhost vagrant]# docker node update --availability drain swarm-node1
swarm-node1
[root@localhost vagrant]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS
6jir7f507ur05lrhtc2xq8plt * swarm-master Ready Active Leader
8nk8i2ekx8kykyzmaab475tnw swarm-node1 Ready Drain
qt9ja2qbmq4n040y60g6xbdlz swarm-node2 Ready Active
[root@localhost vagrant]# docker node update --availability active swarm-node1
swarm-node1
[root@localhost vagrant]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS
6jir7f507ur05lrhtc2xq8plt * swarm-master Ready Active Leader
8nk8i2ekx8kykyzmaab475tnw swarm-node1 Ready Active
qt9ja2qbmq4n040y60g6xbdlz swarm-node2 Ready Active
6. Swarm 中部署服务（nginx）
Docker 1.12版本提供服务的Scaling、health check、滚动升级等功能，并提供了内置的dns、vip机制，实现service的服务发现和负载均衡能力
## 1. 部署服务
### 1.1 创建网络
[root@localhost vagrant]# docker network create -d overlay nginx_net
lj3drxnqik170ta6fr7f4rq5i
[root@localhost vagrant]# docker network ls | grep "nginx_net"
lj3drxnqik17 nginx_net overlay swarm
### 1.2 部署服务
// 就创建了一个具有一个副本（--replicas 1 ）的nginx服务，使用镜像nginx
[root@localhost vagrant]# docker service create --replicas 1 --network nginx_net --name my_nginx -p 80:80 ng
inx
ufxp6pe02z3xggd73ni2jcxay
### 1.3 查看正在运行的服务列表
[root@localhost vagrant]# docker service ls
ID NAME MODE REPLICAS IMAGE
ufxp6pe02z3x my_nginx replicated 1/1 nginx:latest
## 2. 查询Swarm中服务的信息
[root@localhost vagrant]# docker service inspect --pretty my_nginx
ID: ufxp6pe02z3xggd73ni2jcxay
Name: my_nginx
Service Mode: Replicated
Replicas: 1
Placement:
UpdateConfig:
Parallelism: 1
On failure: pause
Max failure ratio: 0
ContainerSpec:
Image: nginx:latest@sha256:06e4235e95299b1d6d595c5ef4c41a9b12641f6683136c18394b858967cd1506
Resources:
Networks: nginx_net
Endpoint Mode: vip
Ports:
PublishedPort 80
Protocol = tcp
TargetPort = 80
## 3. 查看服务在哪个节点上；（在swarm-master节点）
[root@localhost vagrant]# docker service ps my_nginx
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS
t7o6f7hrjd9v my_nginx.1 nginx:latest swarm-master Running Running 6 minutes ago
## 4. 在Swarm中动态扩展服务（scale）
// 容器动态扩展到4个
[root@localhost vagrant]# docker service scale my_nginx=4
my_nginx scaled to 4
[root@localhost vagrant]# docker service ps my_nginx
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS
t7o6f7hrjd9v my_nginx.1 nginx:latest swarm-master Running Running 58 minutes ago
ek7ckkc1g1b4 my_nginx.2 nginx:latest swarm-node2 Running Preparing 5 seconds ago
jdl14u7a8or5 my_nginx.3 nginx:latest swarm-node1 Running Preparing 5 seconds ago
2vflrnohh6h8 my_nginx.4 nginx:latest swarm-node1 Running Preparing 5 seconds ago
7. Swarm 服务（nginx）--- 故障切換
如果一个节点宕机了（即该节点就会从swarm集群中被踢出），则Docker应该会将在该节点运行的容器，调度到其他节点，以满足指定数量的副本保持运行状态。
# 查看服务所在节点
[root@localhost vagrant]# docker service ps my_nginx
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS
t7o6f7hrjd9v my_nginx.1 nginx:latest swarm-master Running Running 58 minutes ago
ek7ckkc1g1b4 my_nginx.2 nginx:latest swarm-node2 Running Preparing 5 seconds ago
jdl14u7a8or5 my_nginx.3 nginx:latest swarm-node1 Running Preparing 5 seconds ago
2vflrnohh6h8 my_nginx.4 nginx:latest swarm-node1 Running Preparing 5 seconds ago
模拟其中一个节点机器关机了
# swarm-node1 节点关闭docker-service
[root@localhost vagrant]# systemctl stop docker查看主节点信息
[root@localhost vagrant]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS
6jir7f507ur05lrhtc2xq8plt * swarm-master Ready Active Leader
8nk8i2ekx8kykyzmaab475tnw swarm-node1 Down Active
qt9ja2qbmq4n040y60g6xbdlz swarm-node2 Ready Active可以看到 swarm-node1 已经关闭了
[root@localhost vagrant]# docker service ps my_nginx
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS
t7o6f7hrjd9v my_nginx.1 nginx:latest swarm-master Running Running about an hour ago
ek7ckkc1g1b4 my_nginx.2 nginx:latest swarm-node2 Running Running 19 minutes ago
buaz7bbd2tjx my_nginx.3 nginx:latest swarm-node2 Running Running 2 minutes ago
jdl14u7a8or5 \_ my_nginx.3 nginx:latest swarm-node1 Shutdown Running 2 minutes ago
5puac9f6ya0j my_nginx.4 nginx:latest swarm-node2 Running Running 2 minutes ago
2vflrnohh6h8 \_ my_nginx.4 nginx:latest swarm-node1 Shutdown Running 2 minutes ago我们可以看到原来 swarm-node1 上的nginx 已经关闭了，而且在其他节点上再创建了等量的 nginx 服务
当我们重新恢复 swarm-node1 节点的 docker 服务（模拟机器恢复了）
[root@localhost vagrant]# docker service ps my_nginx
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PO
RTS
t7o6f7hrjd9v my_nginx.1 nginx:latest swarm-master Running Running about an hour ago
ek7ckkc1g1b4 my_nginx.2 nginx:latest swarm-node2 Running Running 24 minutes ago
buaz7bbd2tjx my_nginx.3 nginx:latest swarm-node2 Running Running 7 minutes ago
jdl14u7a8or5 \_ my_nginx.3 nginx:latest swarm-node1 Shutdown Failed 2 minutes ago "No such container: my_nginx.3…"
5puac9f6ya0j my_nginx.4 nginx:latest swarm-node2 Running Running 7 minutes ago
2vflrnohh6h8 \_ my_nginx.4 nginx:latest swarm-node1 Shutdown Failed 2 minutes ago "No such container: my_nginx.4…"
分配到 swarm-node1 节点的并没恢复
8. 多服务Swarm部署到集群
docker-compose 和 Swarm 结合实现对多服务进行编排
- nginx服务
- visualizer服务
- portainer服务
docker service部署的是单个服务，我们可以使用docker stack进行多服务编排部署
创建 docker-compose.yml
version: "3"
services:
nginx:
image: nginx
ports:
- 8888:80
deploy:
mode: replicated
replicas: 3
visualizer:
image: dockersamples/visualizer
ports:
- "8080:8080"
volumes:
- "/var/run/docker.sock:/var/run/docker.sock"
deploy:
replicas: 1
placement:
constraints: [node.role == manager]
portainer:
image: portainer/portainer
ports:
- "9000:9000"
volumes:
- "/var/run/docker.sock:/var/run/docker.sock"
deploy:
replicas: 1
placement:
constraints: [node.role == manager]通过 docker-compose.yaml 部署服务
[root@localhost testswarm]# docker stack deploy -c docker-compose.yml deploy_deamon
Creating network deploy_deamon_default
Creating service deploy_deamon_visualizer
Creating service deploy_deamon_portainer
Creating service deploy_deamon_nginx
# 查看状态
[root@localhost testswarm]# docker service ls
ID NAME MODE REPLICAS IMAGE
il2vuomkg3by deploy_deamon_portainer replicated 1/1 portainer/portainer:latest
sudtu6byntye deploy_deamon_visualizer replicated 1/1 dockersamples/visualizer:latest
ywtzr8ssersf deploy_deamon_nginx replicated 3/3 nginx:latest