Excellent tool for rapid DevOps

Use Cases and Deployment Scope

Reducing deployment friction is the chief benefit. The advent of the Cloud allowed us to break away from physical servers and rapidly develop our products and services. Docker adds a Saturn 5 rocket to that. It allows developers to easily build very complex application on their local machines much the same way VMs did, but with a fraction of the resources VMs take because of their reduced footprint. This in turn allows sharing prebuilt "containers" with a community be it within an organisation or global in a matter of seconds. Additionally, the days of configuration drift are solved because we all have a common "language" for sharing configurations regardless of the tools used to build them.

Pros and Cons

The best way to eliminate typical dev/testing/prod environment issue.
Simplified the deployment practice for most service systems.
Also contains a large number of pre-built images includes most frequently used services, Nginx, MySQL, etc. Implemented a new way to set up a service with nearly zero impact on the host.

Different running model compares to traditional software development/deployment method
Introduced a little complicated management of networking/storage
Difficult to remote manage/monitoring
Resource sharing between containers or hosts is complex
The docker hub is limited to a free user, can only have one private image

Return on Investment

The dependencies and special steps in the build process are now 100% documented and fully reproducible.
Deployments are easy and the same deployment can be used by any team that deploys Docker containers.
Docker takes out the super complex process required when setting up a new infrastructure. Once everything is ready, deploying everything towards production is vastly simple and streamlined.

Alternatives Considered

Docker is absolutely amazing when it comes to QA environments and setting up developers with the environment they need. For those that are getting started, Kitematic is an amazing interface to use with Docker. We have multiple instances of databases alongside Docker and it's amazing to be able to sandbox these environments.

Likelihood to Recommend

Docker is basically a wrapper for the LXC Container, but it does much more than that. With Docker, you will be able to put everything you need, like a web app, servers or even databases, in a container and then ship it to your production environment. With Docker Compose, you can orchestrate a whole infrastructure based on a Docker container; you just have to describe your setup in a YAML file. It's really infrastructure as code. Furthermore, you can even manage Docker Cluster with Swarm, which will allow you to deploy containers on multiples hosts as if it was only one Docker engine.