swarm集群操作详解
竺鸿骞
承接上篇的集群搭建操作的结束之后，此篇讲解集群的相关操作命令
swarm集群常用操作
- 写在前面:如果集群损坏切在不影响集群业务的下需要重置的
使用docker swarm init --force-new-cluste
节点信息查看
命令位置：集群节点
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active Reachable 18.09.4
hka4wtaqt05hsme3c5ycp2nmp swarm03 Ready Active 18.09.4
筛选查看命令
①只展示id
[root@swarm01 /]# docker node ls -q
2g6592p43eonilcjvivw3ww1f
x64fuzm4y70ov30i5txx327a9
hka4wtaqt05hsme3c5ycp2nmp
[root@swarm01 /]#
②筛选满足条件（支持参数 name id）
[root@swarm01 /]# docker node ls -f name=swarm01
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
修改节点
命令位置：集群节点
节点降低：manager-> worker
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active Reachable 18.09.4
hka4wtaqt05hsme3c5ycp2nmp swarm03 Ready Active 18.09.4
[root@swarm01 /]# docker node demote swarm02
Manager swarm02 demoted in the swarm.
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active 18.09.4
hka4wtaqt05hsme3c5ycp2nmp swarm03 Ready Active 18.09.4
节点升级: worker -> manager
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active 18.09.4
hka4wtaqt05hsme3c5ycp2nmp swarm03 Ready Active 18.09.4
[root@swarm01 /]# docker node promote swarm02
Node swarm02 promoted to a manager in the swarm.
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active Reachable 18.09.4
hka4wtaqt05hsme3c5ycp2nmp swarm03 Ready Active 18.09.4
指定参数修改
--advertise-addr: ip
--listen-addr: ip与端口
--availability: 节点的有效性("active"|"pause"|"drain")
--role：worker manager
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active Reachable 18.09.4
hka4wtaqt05hsme3c5ycp2nmp swarm03 Ready Active 18.09.4
[root@swarm01 /]# docker node update --role worker swarm02
swarm02
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active 18.09.4
hka4wtaqt05hsme3c5ycp2nmp swarm03 Ready Active 18.09.4
移除节点
移除节点完成操作
①命令位置： swarm03(准备移除的节点)
[root@swarm03 ~]# docker swarm leave
Node left the swarm.
[root@swarm03 ~]#
②命令位置：swarm01(控制节点)
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active 18.09.4
hka4wtaqt05hsme3c5ycp2nmp swarm03 Ready Active 18.09.4
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active 18.09.4
hka4wtaqt05hsme3c5ycp2nmp swarm03 Down Active 18.09.4
[root@swarm01 /]# docker node rm swarm03
swarm03
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active 18.09.4
工作节点执行docker node leave之后在集群节点状态变为Down，这样在集群节点执行docker node rm便可以
如果需要强制移除节点，直接执行会报错
[root@swarm01 /]# docker node rm swarm03
Error response from daemon: rpc error: code = FailedPrecondition desc = node 2wo6hxhxa3loms06o3midxhdj is not down and can't be removed
使用强制参数
[root@swarm01 /]# docker node rm -f swarm03
swarm03
[root@swarm01 /]# docker node ls
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
2g6592p43eonilcjvivw3ww1f * swarm01 Ready Active Leader 18.09.4
x64fuzm4y70ov30i5txx327a9 swarm02 Ready Active 18.09.4
[root@swarm01 /]#
部署应用
创建服务
docker service create
–detach , -d: 指定容器运行于前台还是后台，默认为false
–name: 服务名称
–network: 网络连接
–publish , -p: 端口映射
–env , -e: 设置环境变量
–tty , -t: 分配tty设备，该可以支持终端登录
–mount: 文件挂载
–replicas: 指定任务数量
拉取hello-world最新版本镜像作为演示demo
[root@swarm01 usr]# docker service create --replicas 3 -p 443:8080 --name hello \
> --mount type=bind,source=/storage/apiapp_log,destination=/go/src/logs \
> --mount type=bind,source=/root/apiapp/app.conf,destination=/root/app.conf \
> --mount type=bind,source=/root/apiapp/run_application.sh,destination=/root/run_application.sh \
> -e RUN_MODE=Production \
> -td hello-wrold:latest
image hello-wrold:latest could not be accessed on a registry to record
its digest. Each node will access hello-wrold:latest independently,
possibly leading to different nodes running different
versions of the image.
xfmlwf4eo9b27jlhrh889wooy
[root@swarm01 usr]#
查看服务创建情况
[root@swarm01 usr]# docker service ls
ID NAME MODE REPLICAS IMAGE PORTS
xfmlwf4eo9b2 hello replicated 0/3 hello-wrold:latest *:443->8080/tcp
[root@swarm01 usr]#
服务参数修改(这里的镜像因为网速原因用的是最小的hello-world，错误地方暂且忽略)
docker service update [OPTIONS] SERVICE
[root@swarm01 /]# docker service update --replicas 2 hello
hello
overall progress: 0 out of 2 tasks
1/2: preparing [=================================> ]
2/2: preparing [=================================> ]
service update paused: update paused due to failure or early termination of task 4yk09qa7z8rw11517ucmxgxku
针对任务数量修改，支持批量修改的命令为：docker service scale service=replicas。。。 支持批量操作
查看服务
[root@swarm01 ~]# docker service ls （options）
ID NAME MODE REPLICAS IMAGE PORTS
xfmlwf4eo9b2 hello replicated 0/3 hello-wrold:latest *:443->8080/tcp
--filter 同样可以支持参数过滤操作
查看任务列表
docker service ps [OPTIONS] SERVICE [SERVICE...] （支持filter过滤操作）【支持批量操作】
[再次声明，所有展示为了操作命令，运行状态是否正确这里不考虑]
[root@swarm01 /]# docker service ps hello
ID NAME IMAGE NODE DESIRED STATE CURRENT STATE ERROR PORTS
jfpqsqh46n4h hello.1 hello-wrold:latest swarm02 Running Preparing 11 seconds ago
vovkm6h5g4zk \_ hello.1 hello-wrold:latest swarm01 Shutdown Rejected 12 seconds ago "invalid mount config for type??
unuabe9yy49l \_ hello.1 hello-wrold:latest swarm03 Shutdown Rejected 38 seconds ago "invalid mount config for type??
l8apmw48cjiq \_ hello.1 hello-wrold:latest swarm02 Shutdown Rejected about a minute ago "invalid mount config for type??
ue2057am4lps \_ hello.1 hello-wrold:latest swarm01 Shutdown Rejected about a minute ago "invalid mount config for type??
k1i9e990753w hello.2 hello-wrold:latest swarm02 Running Preparing 40 seconds ago
2ezv34jno81a \_ hello.2 hello-wrold:latest swarm03 Shutdown Rejected 41 seconds ago "invalid mount config for type??
k3qwlh6ocukc \_ hello.2 hello-wrold:latest swarm02 Shutdown Rejected about a minute ago "invalid mount config for type??
h9jqy8coh3qd \_ hello.2 hello-wrold:latest swarm03 Shutdown Rejected about a minute ago "invalid mount config for type??
nujear2hj645 \_ hello.2 hello-wrold:latest swarm01 Shutdown Rejected 2 minutes ago "invalid mount config for type??
hyhstj0fngak hello.3 hello-wrold:latest swarm02 Running Preparing 15 seconds ago
xeqe7xgebbd4 \_ hello.3 hello-wrold:latest swarm01 Shutdown Rejected 15 seconds ago "invalid mount config for type??
elmh2v12o16s \_ hello.3 hello-wrold:latest swarm03 Shutdown Rejected 51 seconds ago "invalid mount config for type??
d7u4yxhu70db \_ hello.3 hello-wrold:latest swarm01 Shutdown Rejected about a minute ago "invalid mount config for type??
vcwia4wqthfq \_ hello.3 hello-wrold:latest swarm02 Shutdown Rejected about a minute ago "invalid mount config for type??
查看服务详情
docker service inspect [OPTIONS] SERVICE [SERVICE...] 支持多个操作
[root@swarm01 /]# docker service inspect hello
删除服务
docker service rm SERVICE [SERVICE...]
[root@swarm01 /]# docker service ls
ID NAME MODE REPLICAS IMAGE PORTS
xfmlwf4eo9b2 hello replicated 0/3 hello-wrold:latest *:443->8080/tcp
[root@swarm01 /]# docker service rm hello
hello
[root@swarm01 /]# docker service ls
ID NAME MODE REPLICAS IMAGE PORTS
[root@swarm01 /]#
查看服务日志
docker service logs [OPTIONS] SERVICE|TASK
--follow , -f: 跟随输出
--tail: 显示最新的N条日志
--since: 显示从某个时间开始的所有日志
--timestamps , -t: 显示时间戳
此篇文章建立在dokcer swarm集群搭建完成的基础纸上，可以参见地址：https://zhuanlan.zhihu.com/p/61535537
类似资料：