We already saw how Docker Swarm provides self-healing applied to services. If a replica of a service goes down, it will be rescheduled on one of the healthy nodes. Soon after, the desired number of replicas will be running inside the cluster. We combined that feature with volumes attached to network drives so that the state of the stateful services is persisted.
The time has come to explore how to accomplish the same self-healing features applied to infrastructure. We already know how to create a cluster based on the Docker For AWS template or its equivalent in Azure. If you are not using one of those hosting vendors, we explored the essential features you should implement yourself.
Before we move into infrastructure self-healing, we need to create the cluster we destroyed at the end of the previous chapter. We'll use this opportunity...