Consul in Docker, designed to be self-operating according to the autopilot pattern. This application demonstrates support for configuring the Consul raft so it can be used as a highly-available discovery catalog for other applications using the Autopilot pattern.
- If you don't have it, install the Docker for Windows, docker for Mac, or Docker Toolbox(including
dockeranddocker-compose) on your laptop or other environment. - Create a swarm mode cluster
The option --engine-opt experimental is not mandatory but on docker v1.13 you can run docker service logs consul, to view the logs from all the cluster service.
for i in 1 2 3; do
docker-machine create -d virtualbox --engine-opt experimental swarm-$i
done
eval $(docker-machine env swarm-1)
docker swarm init --advertise-addr $(docker-machine ip swarm-1)
TOKEN=$(docker swarm join-token -q manager)
for i in 2 3; do
eval $(docker-machine env swarm-$i)
docker swarm join --token $TOKEN --advertise-addr $(docker-machine ip swarm-$i) $(docker-machine ip swarm-1):2377
done
- Create a network for consul
docker network create consul-net -d overlay --subnet=172.20.0.0/24
The option --subnet is not mandatory, is just I want to be on different network that the usually 10.x.x.x that docker assign by default.
- Create a service for the swarm
1.1 Using docker-compose v1.10+
docker deploy -c docker-compose.yml consul
1.2 Manually creating the service
docker service create --network=consul-net --name=consul \
-e 'CONSUL_LOCAL_CONFIG={"skip_leave_on_interrupt": true}' \
-e CONSUL_BIND_INTERFACE='eth0' \
-e CONSUL=consul \
-e CONSUL_CHECK_LEADER=true \
--replicas 3 \
--update-delay 10s \
--update-parallelism 1 \
-p 8500:8500 sdelrio/consul
The image use the official consul as base image, so all environment vars can be used (like CONSUL_LOCAL_CONFIG, CONSUL_BIND_INTERFACE).
CONSUL: Name of the service to ask for other consul peers. The image will use docker DNS to find the other peers. Usually this will be the name of the swarm service.CONSUL_CHECK_LEADER: If istruethe logs will show each health check interval if the container is the leader and or container's IP and the leader's IP:
consul_consul.3.l0e0zr114x50@swarm-1 | 2017/01/26 00:11:22 [CP] I'm leader (172.20.0.6)
consul_consul.2.qwx39safki82@swarm-2 | 2017/01/26 00:11:26 [CP] Leader is 172.20.0.6, I'm 172.20.0.3
consul_consul.4.qws5isxw6gpm@swarm-3 | 2017/01/26 00:11:27 [CP] Leader is 172.20.0.6, I'm 172.20.0.4
consul_consul.3.l0e0zr114x50@swarm-1 | 2017/01/26 00:11:32 [CP] I'm leader (172.20.0.6)
consul_consul.2.qwx39safki82@swarm-2 | 2017/01/26 00:11:36 [CP] Leader is 172.20.0.6, I'm 172.20.0.3
consul_consul.4.qws5isxw6gpm@swarm-3 | 2017/01/26 00:11:37 [CP] Leader is 172.20.0.6, I'm 172.20.0.4
consul_consul.3.l0e0zr114x50@swarm-1 | 2017/01/26 00:11:42 [CP] I'm leader (172.20.0.6)
PROTO: Default value tohttp, for leader election, we could set this env var tohttps. So the election will resolve to${PROTO}://${CONSUL}:8500/v1/status/leader.
The entrypoint will execute consul with containerpilot, you can use the command to set your own parameters, by default the command will need 3 replicas:
agent -server -bootstrap-expect 3 -ui -client=0.0.0.0 -retry-interval 5s --log-level warn -disable-host-node-id
Consul generates node IDs from hostname and my conflict when you run two or more consul containers on the same host. Consul (v0.8.1+) has
-disable-host-node-idflag to avoid this issue.
- Oficial consul image: https://hub.docker.com/_/consul/
- Autopilot Pattern with consul: https://github.com/autopilotpattern/consul
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent