Script that create a swarm cluster and deploy a simple service. Swarm is created with Swarm mode of Engine 1.12. Can be created on
- Virtualbox
- Microsoft Azure
- Digitalocean
- Amazon EC2
Note: currently, if deploying on AWS, only EU (Ireland) region is available. Make sure you use a Key Pairs for this region
./swarm.sh [--driver provider]
[--azure-subscription-id azure_subscription_id]
[--amazonec2-access-key ec2_access_key]
[--amazonec2-secret-key ec2_secret_key]
[--amazonec2-security-group ec2_security_group]
[--digitalocean_token]
[-m|--manager nbr_manager]
[-w|--worker nbr_worker]
[-r|--replica nbr_replica]
[-p|--port exposed_port]
[--service_image image_of_the_service_to_deploy]
[--service_port port_exposed_by_the_service_to_deploy]
Several parameters can be provided
- driver used ("azure", "virtualbox", "digitalocean", "amazonec2") (default: "virtualbox")
- number of manager (default: 3)
- number of worker (default: 5)
- number of replicas for the deployed service (lucj/randomcity:1.1) (default: 5)
- port exposed by the cluster (default: 8080)
- azure subscription id (if azure driver selected)
- digitalocean token (if digitalocean driver specified)
- amazon access key, secret key, security group (currently only for EU (Ireland) region) (if amazonec2 driver is specified)
Let's create a swarm cluster with 2 manager and 2 worker nodes locally (with virtualbox) and using service lucj/randomCity Once deployed service will be available on port 8080 (default port)
$ ./swarm.sh --manager 2 --worker 2 --service_image lucj/randomCity --service_port 80
-> about to create a swarm with 2 manager(s) and 2 workers on virtualbox machines
-> creating Docker host for manager 1 (please wait)
-> creating Docker host for manager 2 (please wait)
-> creating Docker host for worker 1 (please wait)
-> creating Docker host for worker 2 (please wait)
-> init swarm
Swarm initialized: current node (99xi3bzlgobxmeff573qitctg) is now a manager.
-> join manager 2 to the swarm
Node f4wocnel60xwfn2z522a645ba accepted in the swarm.
-> join worker 1 to the swarm
This node joined a Swarm as a worker.
-> join worker 2 to the swarm
This node joined a Swarm as a worker.
-> deploy service with 5 replicas with exposed port 8080
-> waiting for service 5ny5u5pmfw75mnomleb34a3kp to be available
... retrying in 2 seconds
... retrying in 2 seconds
... retrying in 2 seconds
... retrying in 2 seconds
... retrying in 2 seconds
... retrying in 2 seconds
... retrying in 2 seconds
... retrying in 2 seconds
... retrying in 2 seconds
-> service available on port 8080 of any node
ID NAME REPLICAS IMAGE COMMAND
5ny5u5pmfw75 city 5/5 lucj/randomcity:1.1
ID NAME SERVICE IMAGE LAST STATE DESIRED STATE NODE
1j157qz7nu4kaqmack4zuwibm city.1 city lucj/randomcity:1.1 Running 20 seconds Running manager1
72y2off8y5f8zp4djzmjdzowg city.2 city lucj/randomcity:1.1 Running 20 seconds Running worker1
efzaweh8lhj9aalrgdhnx26i0 city.3 city lucj/randomcity:1.1 Running 20 seconds Running manager2
1f5ccot3wn3yhrhfbqf6vj5d5 city.4 city lucj/randomcity:1.1 Running 20 seconds Running worker2
f53ummqn8mba0hzy15w08pxj4 city.5 city lucj/randomcity:1.1 Running 20 seconds Running worker2
List all Docker host created
$ docker-machine ls
NAME ACTIVE DRIVER STATE URL SWARM DOCKER ERRORS
manager1 - virtualbox Running tcp://192.168.99.100:2376 v1.12.0-rc2
manager2 - virtualbox Running tcp://192.168.99.101:2376 v1.12.0-rc2
worker1 - virtualbox Running tcp://192.168.99.102:2376 v1.12.0-rc2
worker2 - virtualbox Running tcp://192.168.99.103:2376 v1.12.0-rc2
The test service deployed is a simple http server that returns a message with
- the ip of the container that handled the request
- a random city of the world
Send several requests to the manager1
$ curl 192.168.99.100:8080
{"message":"10.255.0.7 suggests to visit Zebunto"}
$ curl 192.168.99.100:8080
{"message":"10.255.0.8 suggests to visit Areugpip"}
$ curl 192.168.99.100:8080
{"message":"10.255.0.10 suggests to visit Fozbovsav"}
$ curl 192.168.99.100:8080
{"message":"10.255.0.9 suggests to visit Kitunweg"}
$ curl 192.168.99.100:8080
{"message":"10.255.0.11 suggests to visit Aviznuk"}
$ curl 192.168.99.100:8080
{"message":"10.255.0.7 suggests to visit Nedhikmu"}
$ curl 192.168.99.100:8080
{"message":"10.255.0.8 suggests to visit Palmenme"}
Send several requests to the worker2
$ curl http://192.168.99.102:8080
{"message":"10.255.0.8 suggests to visit Wehappap"}
$ curl http://192.168.99.102:8080
{"message":"10.255.0.11 suggests to visit Jocuvdam"}
$ curl http://192.168.99.102:8080
{"message":"10.255.0.12 suggests to visit Suvigenuh"}
$ curl http://192.168.99.102:8080
{"message":"10.255.0.9 suggests to visit Jinonat"}
The requests are dispatched in a round robin fashion to the running containers.
Run 3 managers and 6 workers on Microsoft Azure based on ehazlett/docker-demo image (default image if none specified)
./swarm.sh --driver azure --azure-subscription-id $AZURE_SUBSCRIPTION_ID --manager 3 --worker 6
Run 3 managers and 6 workers on DigitalOcean and use service based on ehazlett/docker-demo image (default image if none specified)
./swarm.sh --driver digitalocean --digitalocean_token $DO_TOKEN --manager 3 --worker 6
Once the service is deployed you got some nice Mobydock :)
Note: beware of the browser cache that prevents the hostname to be updated sometimes
Run 3 managers and 6 workers on AmazonEC2 based on ehazlett/docker-demo image (default image if none specified)
./swarm.sh --driver amazonec2 --amazonec2-access-key $AWS_ACCESS_KEY --amazonec2-secret-key $AWS_SECRET_KEY --amazonec2-security-group default --manager 3 --worker 6
note: make sure the security group provided (default in this example) allow communication between hosts and open the exposed port (8080 by default) to the outside
- Azure deployment with image / size / region selection
- DigitalOcean deployment with image / size / region selection
- Amazon deployment with AMI / instance type / region selection
- Amazon deployment with automatic opening of exposed port in SecurityGroup