• AWS CloudFormation (JSON and YAML)
• Azure Resource Manager (JSON)
• Google Cloud Deployment Manager (YAML)
(By the way, traditional tools and scripting — Bash/perl/cron, etc. — are cer-
tainly still useful.)
If you lean toward programming and development, then Java, Node JS and,
especially, Python are of great value, and used regularly in various cloud ser-
vices. This is relevant for Big Data tools on Hadoop/Spark, and for function as
a service (FaaS) offerings, such as AWS Lambda, Azure Functions and Google
Cloud Functions. These allow the creation and operation of application logic
without the need to provision servers.
Step Three: Learn and Understand DevOps Tools
While the underlying languages that are used to deploy and maintain infra-
structure are JSON and YAML, higher level tools are often used to enhance
deployment. Examples of these are technologies such as Ansible, Chef, Puppet
(configuration management for servers), Terraform and Jenkins (orchestration
of infrastructure as code subcomponents).
Step Four: Consider Containers
Many organizations are using containers to create a layer of portability and
abstracti on between servers and applications. Docker and Kubernetes are the
leading platforms, and the cloud providers are all supporting containers as a
service (CaaS) offerings. Not every organization is moving in this direction, but
this area represents a significant opportunity for many.
Get Experience
Regardless of the guided training you get, nothing substitutes for hands-on
experience. AWS, Azure and Google offer free accounts for your practice — get
one, two or all three. Deploy a LAMP stack from the AWS Marketplace and
practice deployment using code. Learn to use the AWS Command Line Inter-
face (CLI) and Google Cloud Platform’s Cloud Shell. Grab an AWS Quick Start,
deploy a sample reference architecture with a few clicks and dig into how it
works.
Here is the rule of automated cloud infrastructure. Exploring with a web con-
sole is fine, but if you have to go to the console to configure, launch or change
anything in your deployment, you’re doing it wrong. Everything should be done
in code.
Once you have built your platform and have it working, check it into GitHub
and use it as a resume of your work. Go to your employers, tell them what you
have done and demo it for them. Tell them why this is so important to the
health and welfare of the company. If they ignore you, present it to another
company who does understand what cloud will do for their business.
The choice is yours.
WINTER 2018 | THE DOPPLER | 71