Skip to main content

Week 8 - DevOps

Discussion​

Review & Discussion

Front-Ends

DevOps​

DevOps is the intersection, or maybe more accurately - the combination, of development and operations. Let's start by looking at a couple definitions and we'll infer some more tangible meaning.

The Microsoft definition:

DevOps combines development (Dev) and operations (Ops) to unite people, process, and technology in application planning, development, delivery, and operations. DevOps enables coordination and collaboration between formerly siloed roles like development, IT operations, quality engineering, and security.

Source: https://learn.microsoft.com/en-us/devops/what-is-devops

The Amazon definition:

DevOps is the combination of cultural philosophies, practices, and tools that increases an organization’s ability to deliver applications and services at high velocity: evolving and improving products at a faster pace than organizations using traditional software development and infrastructure management processes.

Source https://aws.amazon.com/devops/what-is-devops/

While this definitions are a bit high-level, there is commonality to each of these. First, DevOps involves people, processes, and technology - meaning, it's not only one of these things. Second, DevOps is about increasing organizational efficiency to move faster. Let's begin by looking at some of the practices that enable us to accelerate delivery.

DevOps Practices​

We'll begin by outlining some of the key practices of DevOps.

  • Configuration Management (CM) - Managing the configuration of infrastructure and applications in a consistent and repeatable way. We often do this with tools like Ansible, Chef, or Puppet. It helps us avoid the problem of "snowflake servers" where every server has a slightly different configuration resulting in hard-to-predict behavior.
  • Release Management (RM) - Managing the release of software across environments (e.g. development, staging, production), applying different configurations, and establishing traceability between source code and delivered products.
  • Test Automation - The practice of automating testing to ensure that software is delivered with a high degree of quality. This often includes unit tests, integration tests, and end-to-end tests, but can also involve UI testing, performance and load testing, and security testing.
  • Continuous Integration (CI) - Integrating code changes into a shared baseline frequently (often multiple times per day) and validating those changes with various levels of automated testing. Small, frequent changes with automated validation helps us catch problems early and minimize debugging and rework.
  • Continuous Delivery (CD) - The practice of automating the software delivery process so that software can be reliably released at any time. A step further is Continuous Deployment which is the practice of automatically deploying software to production after it has been validated.
  • Infrastructure as Code (IaC) - The practice of managing infrastructure using code and automation. This goes hand-in-hand with configuration management but involved managing the infrastructure as a flexible, reproducible resource. We often do this with tools like Terraform or CloudFormation.
  • Application Performance Monitoring - Monitoring applications in production to identify issues and opportunities for improvement. This can involve a combination of tools and practices including logging, metrics, and alerts.

Note: These practices were first outlined by Sam Guckenheimer at Microsoft in a course on edX. That course is no longer available so were are unable to link to the original source.

Reading & Async Content​

We don't recommend a ton of reading here. The best way to start learning about this is to begin applying it. This is a topic where we feel concepts don't tend to click until you've gotten hands on with it.

Practice​

Go through each of the above practices for your project and identify how you are going to implement them. Then, start implementing.

We recommend the following order:

  1. Start with CI and automated testing. This may not involve very rigorous testing right away, but should help you get started with a test framework and run some basic sanity checks on your code frequently.
  2. Implement CD along with some level of IaC and CM. Get your software deploying to its production (or production-like) environment automatically. If any configuration needs to happen, introduce that here even if you're not working with multiple environments or configurations yet.
  3. Application Monitoring. Introduce some basic level of monitoring. As a starting point, we suggest rigorous logging and some basic metrics (e.g. a healthcheck endpoint).
  4. After this, it depends on your needs. This is where the needs of your system may dictate whether you've done enough or you need to automate more. You may introduce automated UI testing with tools like Selenium or Cypress or you may want more rigorous system monitoring metrics and automated alerts.