DevOps requires the Cloud – myth or true?

Cloud and DevOps are often described as a perfect match, but the question remains: Is the Cloud necessary for implementing DevOps?

In one word, no. DevOps goes beyond Cloud. How? Well, before we try to understand the slightly complex answer that debunks this myth, let’s start with the basics first.

What is DevOps?

DevOps facilitates and enhances an organization's ability to deliver applications and services much more quickly. In simple terms, DevOps is a combination of cultural philosophies, practices, and tools that ensure faster product development and improvement than organizations which use traditional methods. As one can expect, it’s a no-brainer that organizations with the ability to respond (more) quickly to customer needs will be able to compete more effectively in their markets and serve their customers better.

What about Cloud?

Cloud Computing is the on-demand provisioning of IT resources over the internet on a pay-as-you-go basis. The use of Cloud technology services, such as computing power, storage, and database, is an alternative to investing in, owning, and maintaining physical data centers and servers, with several Cloud providers to choose from. Here are some of the benefits that the Cloud provides:

• Faster time-to-market
• Scalability & flexibility 
• Reduced costs
• Enhanced security 
• Disaster recovery 
• Better collaboration

How should you select a deployment model?

Before we dive into the dynamics of the Cloud and DevOps relationship, let's explore how and why any DevOps team selects a specific deployment model over another. We'd like to explain this with an example.

This example describes a situation where two teams are working on two different projects for their clients. Let's call them, Team A and Team B, for Client A and Client B, respectively. For Client A, Team A is creating a small-scale application with three services to meet the requirements to cater to Client A’s customers. These services expect limited traffic but look forward to keeping user data confidential and under their scope.

On the other hand, Team B is working with Client B to launch an e-commerce website, where they expect traffic to fluctuate, depending on the holiday season. While Client B values user data, they are willing to store it elsewhere.

Based on their extensive analysis, Team A's DevOps team recommends on-premise infrastructure for their client. This would enable them to have full access to the user data and they shall be able to monitor it closely. In contrast, Team B's DevOps team advises Cloud infrastructure for their client, so that in case of any change in the requirements, they can easily scale the infrastructure up or down.

It is worth noting that two DevOps teams of the same company have recommended completely different deployment models (strategies) to their clients, while considering factors such as scaling, costing, and data privacy. Clients have their own requirements for infrastructure and have varying budgets, which means they always prefer to have the best-fit deployment model that caters to their requirements in the best way possible.

But how do Deployment Models make a difference?

On-premise, cloud, or hybrid – how do DevOps deployment models/strategy make a difference?

As a follow-up to our previous question, no matter what type of deployment model is chosen - whether it is an on-premise model, a Cloud model, or a hybrid model - the DevOps methodology remains the same.

Any DevOps team strives to reduce the software development life cycle and deliver quality software in the long run. To fulfill this objective, their effort spans across three phases: Continuous Integration, Continuous Deployment, and Continuous Delivery. These goals can be achieved through tools like Jenkins, Azure DevOps, GitHub Action, Kubernetes, and Octopus.

Continuous Integration entails the creation of executable files for deployment at customers' machines or servers, based on the codebase. Once the files have been created, it is the responsibility of Continuous Delivery to automate the delivery of the latest changes into a system. For Continuous Integration and Continuous Delivery, the processes remain the same for all projects, regardless of the deployment model in use. However, Continuous Deployment takes a different route.

Has the right deployment model been selected or not? This can have a considerable impact on the DevOps process associated with Continuous Deployment. In this phase, the integration between the executable file and the deployment model is set up, and at this point, the executable file is deployed by using the selected deployment model. All deployment models have different setups, requirements, management, and strategies, which must be considered and handled accordingly, before deploying an executable file.

What perks does Cloud have for DevOps?

As far as I am concerned, it is ideal to have a DevOps approach, combined with a Cloud deployment model. Cloud technology and DevOps can simultaneously provide a wide array of DevOps features and functionalities. And with time, the benefits of DevOps would only grow in future. If we can optimally leverage the potential of DevOps and the Cloud, we can reduce the time required to develop software and respond to customer feedback. This will also save a lot of development cost.

Conclusion

Having learned that DevOps isn't just for Cloud platforms, you might now have also realized  that the myth doesn't revolve just around Cloud and DevOps, but that a deployment model is the key to understanding how DevOps and Cloud interact. Although there are many benefits of DevOps for Cloud-based businesses, you don't have to use it exclusively with the Cloud. You can use it for any deployment model, even if it differs from Cloud-based deployment.