Getting on the Right Side of Operations — A Developer’s Guide

Fast. Change. New. These describe the business mandate, purpose and imperatives for developers. When stable, reliable and secure describe the business mandate of Operations, it’s no wonder they’re said to be like oil and water, working in silos with hard barriers between their purposes and needs.

Developers will have come across these common complaints from Ops:

• They think we don’t understand Devs’ work.
• We’re the ones getting calls to fix your software at all hours!
• You don’t understand or care about what we do; if you did, you would get why we shouldn’t move as fast as you want to.
• The system must stay up. Everything else comes after. What don’t you get about that?

Uniting these silos so that together Dev and Ops can deliver applications and services means a better understanding of each other’s needs. And beyond the softer side of better communication and feedback loops, it means sharing practices and tools. So, how can Dev and Ops reach beyond the division of labour to unity of purpose?

First, developers can ask themselves these questions:

• Do I have an understanding of real-time operations and behaviour?
• Do I, and be honest, want to throw my solution over the wall and walk away when it is complete?
• Do I spend time coding for automation, reliability, security and compliance, as well as working on features?
• Do I own how my application behaves in production? 

Developers who appreciate that Operations is responsible for all layers of the application stack, including its stability, security and recoverability can ease Ops’ burdens in a number of ways.

How to mix oil and water

A culture of collective ownership is fostered when Dev and Ops use the same tools, such as configuration templates and data monitoring, work tracking, review and deployment tools. Adding automation is key. Self-help access for Ops services means less fire-fighting for them and streamlines the change process of testing and then moving applications into production. Developers get quick self-access to resources and Ops’ administrative tasks, like provisioning, are easier. Above all, automated security checks are a win for everyone.

Performing automated security audits early in the development process goes a long way to make sure there are no vulnerabilities in your code that can be exploited by hackers. Automated security audits reduce risk for the business, reduce the backlog of checks that Ops needs to run, get rid of repetitive tasks, and bring a shared focus to development, operations and security teams.

A good example is the cloud architecture based on a software-defined data centre that automates application provisioning in a VMware private cloud. It delivers a complete application stack with fully integrated and tested application environments for development and testing engineers. It supports developer choice in application programming interface (API) and graphical user interface (GUI) access to resources. Everything is managed with and defined as code, including policies and blueprints. Log aggregation and analytics are enabled with capabilities for log searching and root cause analysis. 

Benefits span the need for speed for developers, stability for Operations and lower costs for the business—reduced provisioning time, higher developer productivity and reduced infrastructure costs. It also means that Operations, security and development are working in tandem.