DevOps, DevSecOps, and DataOps, it is all about the “Ops” in 2020

This article will give you a small introduction to DevOps, DevSecOps, and DataOps.Let’s break the terms down to details:

· DevOps = Development and Operations.

· DevSecOps = Development, Security, and Operations.

· DataOps = Data and Operations.

Usually, these terms are associated with approaches to Software Development and IT-Operation-, projects, and teams solely. But from what I have understood, it has a more significant potential to impact the way of doing business in the 20th century. So why is it so important to know about concepts associating with a software development approach?

Because today more or less, everything happens with, through, or on software. Let’s face it; software runs everywhere. From accounting, sales, customer service, marketing, finance, production, supply, and so on. The ones making all these coveted software and software companies have unlocked something we all have to investigate and to know about, going into 2020.

DevOps

DevOps brings a new era to company culture, use of tools, and practices. That equips the organization’s ability to serve the customer better.

As part of a DevOps approach, software development and IT- operations departments/teams are no longer in isolated “silos.” They work as one team in a continuous flow, where the engineers work across the entire lifecycle of the product, from development, test, release/production, and operation. This equips engineers with a set of skills and a broader view, not limited to one function or domain only.

DevOps process for a team’s production flow (development, test, and production environment), will following similar iterations as shown in the model of an infinity loop below:

Figure 1. Model of a typical DevOps infinity loop (By Anders Eide)

DevOps integrates the earlier silos in Software development and IT-operations practices, to better create value for its customers with a quicker response to its needs. By speeding up the build lifecycle (known as release engineering) to reduces time, time of deployment, resolving issues, time to market and the market response. At the end of the cycle “Learn” feeds back to the beginning “Plan”, and the process iterates indefinitely, based on customer's feedback, interactions, and data (continuous improvement / continuous development).

DevSecOps

When IT-security and quality assurance are well established as part of all the steps; from end-to-end in the infinity loop, this is referred to as DevSecOps.

DevOps / DevSecOps approach is not only focusing on integration and tight collaboration between department/ team members ( Softeare Development, IT-Operation and IT- Security), but it also relies on automation. Automation to speed up delivery processes, so the product can reach the market fast and the market response be tracked; then acted on, as soon as possible.

The practice collecting, processing, interpret, and act seamlessly on market data after a software product is released, is key to DevOps. It gives the company a cutting edge on understanding what the customers prefers even before the consumers knows it.

DataOps:

It takes good data quality, interpretations, and processing for an organization to respond and act precisely on its data; to achieve great impact. In the evolution of DevOps, DataOps seeks to reduce the end-to-end cycle time of data flows in a Data science/data analytics environment. From the raw data access to ideas, creation of models, charts, and graphs, it relies on innovation and tools.

Figur 2. Plain illustration of the evolution to DataOps.

The data flow in a Data Science environment enters a data pipeline, progresses through a series of calculations, and exits as a form of a report, model and views (see figure3.). The job of a data pipeline is to quality check the data and to manage the efficiency, constraints, and uptime of the environment. The combinations of pipelines and calculations are often called a “DataFactory”. DataOps is all about orchestrating, monitor and manage the data factory.

Figure 3. DataOps illustration with one pipeline.

In comparison to pure DevOps, DataOps facilitate an environment for Data Scientists to innovate. The lifecycle of DataOps shares the same iterative processes like DevOps, but it contains two active pipelines. Figure 3 illustrates one of the two pipelines. The other pipeline is the one handling the main pipeline update. Update of new measures and functionality. See figure 4 for illustration.

Figure 4. Illustration of the main production pipeline and the innovation pipeline.

Figure 3. DataOps illustration with value and innovation pipeline.

This article gives just a little introduction to some of the evolution, aspects, and concepts of DevOps. I hope this gives you a tasteful sneak peek to further check out, how and what DevOps can do for your organization and/or team. I will further come back and write a more in-depth article on this topic; from a business perspective and a technical one.