Choosing a Deployment Strategy: A CTO’s Guide

This article provides a basic guide on selecting your deployment strategy depending upon your business and engineering needs. We will discuss deployment strategies and practices that will guide your engineering team to succeed with their production deployments.

As you are aware we live in the “DevSecOps” era where systems are flexible and scalable. We offer a more adaptive functionality leveraging a modern technology stack, microservices architecture, and cloud infrastructure.

Each organization has a unique requirement, and they deploy their application features in production per day, per month, or quarter. No doubt that your customers like new features but a higher frequency of deployment can also add the risk of downtime. Choosing the right deployment strategy is very important to minimize the risks without customers noticing that something has changed in your production environment.

Before you invest in an expensive one-click deployment strategy, ask yourself three questions:

1. What are your business service level agreements?

• Will your deployment affect your customer’s revenue?
• Will you see an increase in customer churn due to higher downtimes?
• What is your customer’s SLA for their customers?

2. Do you have a sufficient infrastructure budget?

• Do you have a budget for 24X7 infrastructure?
• Have you invested in automation and tooling?
• Have you invested in the right technical skills to support complex deployments?

3. Do you have a deployment testing strategy?

• How do you plan to release and test new features?
• Are you releasing new features to a subset of customers and why?
• Does your customer even care for frequent releases?

Once you know your business and technical outcomes, start researching the deployment patterns including pros and cons, and then draft your strategy or a mix of strategies. There is no one-size-fits-all approach.

Which Strategy Fits Your Needs?

There are multiple ways to deploy a new version of an application, and it depends on your business needs and IT budget. We can use a variety of deployment strategies as listed below, and each strategy has its pros and cons. Every strategy has its trade-offs but if you select the right strategy, you’ll have zero downtime.

Diagram Reference: Google Cloud Release Engineering

Some strategies may be costly to use in your cloud Dev, Test, or Staging environments, whereas the costs are worth it in a Production deployment. Fully understanding the approach and alternative options is very important.

Key things to consider

• Reduce manual and repetitive tasks by investing in automation and tooling.
• Enable continuous monitoring of your deployment pipeline.
• Measure your Service Level Agreements (SLA), Service Level Objectives (SLO), and Service Level Indicators (SLI)
• Frequently deploy and if it hurts then do more !!

Conclusion

Depending on your business and engineering needs, you can deploy and release your application in several ways. Each deployment or testing pattern has its advantages and disadvantages. You should be following the basic principles of release engineering where you have reproducible and automated builds with automated testing and deployments.

The best choice comes down to the needs and constraints of your business where you can deploy the features smoothly. The end-users should not be impacted thereby minimizing the downtime and last-minute failures.

Your deployment mantra: “Release early, Release often, Release in small deployments, and listen to the voice of your customers”