生産環境部署swarm叢集及 swarm-overlay-portainer(webui) 實戰應用

參考文檔：排程過濾器：filter https://docs.docker.com/swarm/scheduler/filter/#how-to-write-filter-expressions 排程政策:strategy https://docs.docker.com/swarm/scheduler/strategy/#spread-strategy-example 部署生産環境swarm叢集 https://docs.docker.com/swarm/install-manual/#step-5-create-swarm-cluster WEBUI portainer https://portainer.readthedocs.io/en/latest/deployment.html Overlay 網絡搭建 http://note.youdao.com/noteshare?id=e02172789830ac387b8ad0216e984b8a

一、安裝步驟 1. Set up a discovery backend : 10 上啟動consul docker run --restart=always -d -p 8500:8500 -h consul --name consul progrium/consul \ -server -bootstrap -advertise 172.19.9.10

2. Start the Swarm manager : 20/30 上啟動manage01 manage02 docker run -d --restart=always --name swarm-manage01 -p 4000:4000 172.19.9.10:5000/swarm \ manage -H :4000 --replication --advertise 172.19.9.20:4000 consul://172.19.9.10:8500

docker run -d --restart=always --name=swarm-manage01 -p 4000:4000 172.19.9.10:5000/swarm \ manage -H :4000 --replication --advertise 172.19.9.30:4000 consul://172.19.9.10:8500

3. Connect to node01 and node02 in turn and join them to the cluster : 40/50 上啟動node01 node02 docker run -d --restart=always --name=swarm-node01 172.19.9.10:5000/swarm \ join --advertise=172.19.9.40:2375 consul://172.19.9.10:8500

docker run -d --restart=always --name=swarm-node02 172.19.9.10:5000/swarm \ join --advertise=172.19.9.50:2375 consul://172.19.9.10:8500

4. 添加--label标簽 40 #docker daemon --label region=huilongguan vim /usr/lib/systemd/system/docker.service ExecStart=/usr/bin/dockerd --graph=/data1/docker -H unix:///var/run/docker.sock -H tcp://0.0.0.0:2375 --cluster-store=consul://172.19.9.10:8500/network --cluster-advertise=em1:2375 --insecure-registry=172.19.9.10:5000 --label region=huilongguan 50 #docker daemon --label region=shangdi vim /usr/lib/systemd/system/docker.service ExecStart= ...... --label region=shangdi

二、測試結果 Node filters Use a constraint filter

docker -H :4000 run -d --net=internaloverlay --name swarm-test-redis-0 -e constraint:node==server3.riskdetection 172.19.9.10:5000/redis docker -H :4000 run -d --net=internaloverlay --name swarm-test-redis-1 -e constraint:node==server4.riskdetection 172.19.9.10:5000/redis docker -H :4000 exec -ti swarm-test-redis-1 ping swarm-test-redis-0 PING swarm-test-redis-0 (10.10.10.10): 56 data bytes 64 bytes from 10.10.10.10: icmp_seq=0 ttl=64 time=1.261 ms 64 bytes from 10.10.10.10: icmp_seq=1 ttl=64 time=0.314 ms 64 bytes from 10.10.10.10: icmp_seq=2 ttl=64 time=0.292 ms 64 bytes from 10.10.10.10: icmp_seq=3 ttl=64 time=0.293 ms 64 bytes from 10.10.10.10: icmp_seq=4 ttl=64 time=0.284 ms 64 bytes from 10.10.10.10: icmp_seq=5 ttl=64 time=0.273 ms 64 bytes from 10.10.10.10: icmp_seq=6 ttl=64 time=0.259 ms

docker -H :4000 run -d --net=internaloverlay --name swarm-test-redis-hlg -e constraint:region==huilongguan 172.19.9.10:5000/redis docker -H :4000 run -d --net=internaloverlay --name swarm-test-redis-sd -e constraint:region==shangdi 172.19.9.10:5000/redis docker -H :4000 exec -ti swarm-test-redis-sd ping swarm-test-redis-hlg PING swarm-test-redis-hlg (10.10.10.9): 56 data bytes 64 bytes from 10.10.10.9: icmp_seq=0 ttl=64 time=0.889 ms 64 bytes from 10.10.10.9: icmp_seq=1 ttl=64 time=0.268 ms 64 bytes from 10.10.10.9: icmp_seq=2 ttl=64 time=0.297 ms 64 bytes from 10.10.10.9: icmp_seq=3 ttl=64 time=0.254 ms 64 bytes from 10.10.10.9: icmp_seq=4 ttl=64 time=0.270 ms 64 bytes from 10.10.10.9: icmp_seq=5 ttl=64 time=0.283 ms 64 bytes from 10.10.10.9: icmp_seq=6 ttl=64 time=0.277 ms

docker -H :4000 ps | grep swarm-test ea8cf21ae46f 172.19.9.10:5000/redis "docker-entrypoint.sh" 40 minutes ago Up 40 minutes server3.riskdetection/swarm-test-redis-hlg c93c6054f5aa 172.19.9.10:5000/redis "docker-entrypoint.sh" 42 minutes ago Up 42 minutes server4.riskdetection/swarm-test-redis-sd 24a092af0e45 172.19.9.10:5000/redis "docker-entrypoint.sh" About an hour ago Up 40 minutes server3.riskdetection/swarm-test-redis-0 eda41c6d1b86 172.19.9.10:5000/redis "docker-entrypoint.sh" About an hour ago Up 41 minutes server4.riskdetection/swarm-test-redis-1 小結：通過overlay 不同主機間容器實作了互聯，根據node,region分别将對應的容器分發到相應的node節點上。

三、注意： Node filters除了 constraint 外，還有以下兩種filter Use the health filter The node health filter prevents the scheduler from running containers on unhealthy nodes. A node is considered unhealthy if the node is down or it can’t communicate with the cluster store.

Use the containerslots filter --label containerslots=3 Swarm will run up to 3 containers at this node, if all nodes are “full”, an error is thrown indicating no suitable node can be found. If the value is not castable to an integer number or is not present, there will be no limit on container number.

Container filters affinity dependency port

四、安裝管理界面WEBUI docker run -d --restart=always --name portainer -p 9000:9000 portainer/portainer -H tcp://172.19.9.20:4000 http://172.19.9.10:9000/ 設定密碼,登陸 1.Cluster info