Docker Swarm for AWS Guide

This container guides the cluster instances through its lifecycle performing maintenance tasks

Example project: Terraform docker-swarm

Description

This container performs several maintenance and cleanup operations for a Docker Swarm node running on AWS.

Tasks

• Cleanup (Manager Nodes) every 5 minutes Removes downed/downscaled nodes from the swarm based on events on a SQS queue.

• Refresh (Manager Nodes) every 4 minutes Updates the DynamodDB table with the current Docker Swarm primary Manager.

• Vacuum (All Nodes) every day between 00 and 01 Runs docker system prune --force to remove all dangling resources.

• Watcher (All Nodes) every minute Handles AutoScalingGroup Instance shutdown by reading events for the node on a SQS queue and notifying the ASG once we are done exiting the swarm.

Usage

Paramaters

Parameter	Example	Description
DYNAMODB_TABLE	-	DynamodDB table ID
NODE_TYPE	worker / manager	Role of the node we are running on
REGION	eu-central-1	AWS Region ID
VPC_ID	vpc-123123	AWS VPC ID
SWARM_QUEUE	-	SQS Queue ID for swarm control
CLEANUP_QUEUE	-	SQS Queue ID for cleaup actions
RUN_VACUUM	yes / no	Enable or Disable Vacuum actions

Example

docker run -d \
  --name=guide-aws \
  --restart=always \
  -e DYNAMODB_TABLE=$DYNAMODB_TABLE \
  -e NODE_TYPE=$NODE_TYPE \
  -e REGION=$AWS_REGION \
  -e VPC_ID=$VPC_ID \
  -e SWARM_QUEUE="$SWARM_QUEUE" \
  -e CLEANUP_QUEUE="$CLEANUP_QUEUE" \
  -e RUN_VACUUM=$RUN_VACUUM \
  -v /var/run/docker.sock:/var/run/docker.sock \
  -v /usr/bin/docker:/usr/bin/docker \
  pecigonzalo/docker-guide-aws