Skip to content
This repository has been archived by the owner on Apr 16, 2024. It is now read-only.

thojkooi/terraform-digitalocean-docker-swarm-mode

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

33 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Terraform DigitalOcean Docker Swarm mode

Terraform module to provision a Docker Swarm mode cluster in a single availability zone on DigitalOcean, using a private network.

CircleCI

Requirements

  • Terraform >= 0.11.7
  • Digitalocean account / API token with write access
  • SSH Keys added to your DigitalOcean account
  • jq

Usage

module "swarm-cluster" {
  source           = "thojkooi/docker-swarm-mode/digitalocean"
  version          = "1.0.0"
  domain           = "do.example.com"
  total_managers   = 3
  total_workers    = 2
  manager_ssh_keys = [1234, 1235, ...]
  worker_ssh_keys  = [1234, 1235, ...]

  providers {}
}

SSH Key

Terraform uses an SSH key to connect to the created droplets in order to issue docker swarm join commands. By default this uses ~/.ssh/id_rsa. If you wish to use a different key, you can modify this using the variable provision_ssh_key. You also need to ensure the public key is added to your DigitalOcean account and it's listed in both the manager_ssh_keys and worker_ssh_keys lists.

Exposing the Docker API

You can expose the Docker API to interact with the cluster remotely. This is done by providing a certificate and private key. See the Docker TLS example for information on how to create these.

module "swarm_mode_cluster" {
  source           = "thojkooi/docker-swarm-mode/digitalocean"
  version          = "1.0.0"
  domain           = "do.example.com"
  total_managers   = 3
  total_workers    = 2
  manager_ssh_keys = [1234, 1235, ...]
  worker_ssh_keys  = [1234, 1235, ...]

  remote_api_ca          = "${path.module}/certs/ca.pem"
  remote_api_certificate = "${path.module}/certs/server.pem"
  remote_api_key         = "${path.module}/certs/server-key.pem"

  manager_size = "s-2vcpu-4gb"
  worker_size  = "s-1vcpu-1gb"
  manager_tags = ["${digitalocean_tag.cluster.id}", "${digitalocean_tag.manager.id}"]
  worker_tags  = ["${digitalocean_tag.cluster.id}", "${digitalocean_tag.worker.id}"]
  providers = {}
}

Note that for this to work, you need to open the Docker remote API port in both iptables (not necessary with default images) and the DigitalOcean cloud firewall.

Notes

Installing Docker

It module does not install Docker - this is left up to the user of this module. The default image used comes with Docker CE pre-installed. It's encouraged to provide your own image or use configuration management tooling to install Docker.

You can also install Docker using user data. See the user-data example.

This module has been tested with Docker CE v18.06 and later. Earlier versions should work (v1.13 and up), but have not been tested.

Supported OS

This module has been tested with Ubuntu Docker (docker-18-04), CoreOS, and CentOS 7.4 provided by DigitalOcean, but it should work with other distributions as well, as long as Docker and sudo packages are installed.

Ports & Firewall

Ensure the following ports are open on the local firewall;

Port Description Note
2377/TCP cluster management communications Cluster
7946/TCP Container network discovery Cluster
7946/UDP Container network discovery Cluster
4789/UDP Container overlay network Cluster
2376/TCP Docker Remote API Optionally, when exposing the Docker Remote API

For example, when using the Docker images provided by DigitalOcean, run the following command:

ufw allow 2377
ufw allow 7946
ufw allow 7946/udp
ufw allow 4789/udp

Cloud Firewall

Also set up firewall rules on DigitalOcean for the cluster, to ensure only cluster members can access the internal Swarm ports. You can use the digitalocean-docker-swarm-firewall module for this. Look in the firewall examples directory for inspiration on how to do this.

Examples

For examples, see the examples directory.

Swarm mode set-up

Manager nodes

First a single Swarm mode manager is provisioned. This is the initial leader node. If you have additional manager nodes, these will be provisioned after this step. Once the manager nodes have been provisioned, Terraform will initialize the Swarm on the first manager node and retrieve the join tokens. It will then have all the managers join the cluster.

If the cluster is already up and running, Terraform will check with the first leader node to refresh the join tokens. It will join any additional manager nodes that are provisioned to the Swarm mode cluster.

Access the API

To expose the Swarm mode API in HA, create a load balancer and forward tcp traffic to port 2376. Ensure you expose the docker remote API using certificates when doing this. Alternatively you can do DNS round-robin load balancing.

When you do not wish to expose your Docker API, you can use SSH to connect to one of the manager nodes and access the Docker API through this.

Worker nodes

Worker nodes should be used to run the Docker Swarm mode Services. By default, 2 worker nodes are provisioned. Set the number of desired worker nodes using the following variable: total_workers.

License

MIT © Thomas Kooi