The FinOps guide to DevOps | Part 1: Cloud infrastructure concepts

The rise of cloud computing has revolutionized the way organizations deploy and manage applications. This led to the emergence of DevOps and FinOps as critical disciplines within modern IT environments. DevOps aims to streamline the development lifecycle and enable continuous, high-quality code delivery.

To support that, DevOps engineers often own the cloud infrastructure – where cloud costs are generated. This makes common language between FinOps and DevOps crucial for effective cloud cost management. In this article, we explore some key DevOps concepts in relation to the cloud infrastructure. We’ll touch on their impact on cloud costs and consumption, and how FinOps can help optimize them.

Continuous Integration and Continuous Deployment (CI/CD)

CI/CD is a cornerstone of DevOps practices, enabling teams to integrate code into a shared repository frequently (Continuous Integration) and automatically release the integrated code to production (Continuous Deployment). This process minimizes the time to bring new features and fixes to the market, aka Time-to-Market, or TTM.

FinOps relevance: CI/CD can increase cloud resource consumption as more environments are spun up in parallel for testing and deployment. FinOps practitioners can work with DevOps teams to ensure that these resources are scaled appropriately, turned off when not in use, and that the right types of resources are chosen for the job to control costs. This can involve using spot instances or discounted commitments to run development and testing jobs.

Microservices and containers

Microservices architecture breaks down applications into smaller, independently deployable services. This means that changes to a single service can be deployed without having to redeploy the entire application – which allows shorter, more resource-efficient Dev, QA, and deployment cycles.

Containers package a chunk of code that constitutes a cloud-native application – usually a microservice – with its dependencies (runtime, system tools, libraries, and settings). Each container can run independently on any compute environment and be started and stopped quickly. This allows fast scaling of applications in response to demand. It also makes more efficient use of cloud infrastructure than monolith (single-VM, non-containerized) apps, allowing for more apps to run on a single virtual machine, or a given set of servers.

FinOps relevance: These technologies may improve resource efficiency. However, they present immense complexity in tracking and managing resource costs. This is because multiple services share the same infrastructure – all with dynamic, fluctuating compute and storage demands. FinOps can help by implementing a good tagging strategy that will support accurate allocation of resource consumption and cost. Then, it will be a joint FinOps-DevOps effort to automate cloud cost management.

Autoscaling

Autoscaling refers to automatically provisioning and adjusting compute resources based on workload (application or job) demands. It ensures that applications have the right amount of resources without manual intervention.

FinOps relevance: Autoscaling can significantly save costs by scaling down resources during low demand. FinOps can monitor patterns and set appropriate thresholds to prevent over-provisioning and avoid waste. These will be configured by the DevOps team into the container orchestrator (e.g. Kubernetes) and/or the autoscaler that works on top of it.

Lastly, they can advocate for autoscaling solutions that include techniques for eliminating idle resources. To learn the top 4 techniques, watch this 3-min. video.

Infrastructure as Code (IaC)

IaC involves managing and provisioning infrastructure through machine-readable definition files, rather than by manually configuring hardware or using interactive user interface (UI) configuration tools. IaC allows for the automation of provisioning and de-provisioning of resources. It therefore ensures that the infrastructure is provisioned consistently, reducing the risk of human error and the associated costs. This consistency also helps with forecasting and budgeting, as it reduces the variability in infrastructure setup.

FinOps relevance: IaC can lead to cloud cost savings if the right autoscaling policies are configured into it. FinOps can leverage IaC to enforce its policies regarding resource tagging, provisioning, scheduling, and cost control. This will help ensure resources are mostly sized properly, shut down when no longer needed, and their consumption is allocated to the right cost center. It is common to use Policy as Code (PaC) to set and enforce such policies in IaC.

Monitoring and logging

DevOps relies on extensive monitoring and logging to track the performance and health of containers, services, applications, and infrastructure, enabling proactive maintenance and troubleshooting. DevOps monitoring is performance-focused and usually does not track or analyze costs – for which FinOps practitioners need their dedicated tooling.

FinOps relevance: Monitoring and logging generates significant amounts of data, which incurs costs for storage and analysis. FinOps should work with DevOps to establish the right balance between the depth of monitoring required for operational efficiency and the cost of data retention and analysis. In addition, FinOps can purchase discounted commitments for hyperscaler storage and analysis services, to optimize rates for the needed capacity.

Conclusion

Nearly any major DevOps practice has the potential to streamline cloud usage and costs. While DevOps teams own the cloud infrastructure’s configurations, FinOps has the knowledge and wider business perspective to offer policies and best practices. By befriending DevOps concepts, FinOps practitioners can ask the right questions about the business’ cloud operations – and understand the answers. This common language is key to an effective collaboration, in which organizations find out what generates excessive cloud costs, and how to mitigate them in the most efficient way.

Ready to test your new knowledge? Click here to learn how you can automate and optimize your container infrastructures.