Using Spot Instances To Maximize Cloud Cost Efficiency And Performance: OpsMx's Experience

At a glance:

Challenge: Managing infrastructure costs for dev, test, and demo environments; needing dynamic provisioning for Spinnaker/Argo clusters; and facing limitations with existing over-provisioned or inelastic clusters.

Rackspace Spot Instance solution: Rackspace Spot enabled greater use of spot instances while managing availability requirements; provided Kubernetes control plane with minimal operational overhead; and offered high reliability with fully managed Kubernetes cluster infrastructure.

Business Outcomes

  • Achieved ~83% reduction in infrastructure costs for test/demo workloads.
  • Improved cluster provisioning time from hours to minutes.
  • Streamlined infrastructure management with built-in monitoring and auto-healing.
  • Enabled scalable usage without vendor lock-in concerns.

About OpsMx

Founded in Silicon Valley in 2017, OpsMx is a leader in secure software delivery and Application Security Posture Management (ASPM). The company helps enterprises accelerate DevSecOps by automating and governing continuous delivery pipelines using tools like Spinnaker, Argo CD, and Jenkins.

OpsMx serves global customers across banking, healthcare, and technology sectors with flexible deployment models—on-prem, SaaS, and hybrid. Its solutions offer AI-driven risk assessment, compliance automation, and software supply chain security, enabling faster, safer software delivery at scale.

The Challenge: Balancing infrastructure costs with dynamic provisioning needs

OpsMx provides modern, CI/CD infrastructure software to leading enterprises.  Their own globally distributed team of software engineers were deeply involved in cloud technologies and cloud-native software development.  However, they were facing a challenge with the cost of their cloud infrastructure.

To solve this, OpsMx had previously tried to optimize cloud costs by using AWS EC2 Spot Instances.  However, like other AWS users, OpsMx found that AWS Spot Instances were only marginally cheaper than savings plans based prices, and therefore they found it hard to justify the additional risk and complexity of Spot Instance pre-emption for these marginal savings.  Eventually, OpsMx switched to using Google Cloud’s GKE which was incrementally better, but still quite expensive vs their desired cost of infrastructure.

OpsMx desired a more efficient way to reduce the cost of their cloud infrastructure, and identified 3 challenges:

  • Managing infrastructure costs for dev, test, and demo environments.
  • Needing quick and easy, dynamic provisioning for Spinnaker/Argo clusters.
  • Existing solutions were either over-provisioned or lacked elasticity.

The Solution: Fully managed Spot Instances with Rackspace Spot 


OpsMx was an early user of Rackspace Spot, since early 2024, when the product was first announced.  Their use of Spot Instances has grown significantly since then, and they have managed to achieve their desired goals of lowering cloud costs while also delivering a better provisioning experience to their users.  

From their early days starting with just 1 cluster in Rackspace Spot, OpsMx today uses 10s of Kubernetes clusters running a variety of workloads, from development to QA to staging and production:

  • Rackspace Spot was very easy to onboard and integrate with existing CI/CD pipelines.
  • The solution was first used extensively for internal demos; then dev environments, and testing Spinnaker/Argo upgrades which end up being quite infrastructure intensive.
  • Setting up resilient clusters using Spot Instances takes just a few minutes using the Spot Terraform provider, UI or API, and OpsMx uses all of these interfaces.

What OpsMx experienced with Rackspace Spot:

  • Seamless integration with spot instances for maximum cost savings.
  • Spot Instance transparency via price points and capacity availability allowed OpsMx to use Spot Instances more aggressively vs more expensive on-demand instances or savings plans.  With other cloud providers, OpsMx found that fear of pre-emption and lack of transparency into spot instance pricing and availability, would result in operations teams having to make conservative decisions which ended up being much more expensive.
  • Fully managed Kubernetes control plane with zero operational overhead. OpsMx teams never had to worry about deploying, monitoring, troubleshooting or upgrading. Kubernetes control plane.
  • High reliability with fully managed infrastructure and simplified cluster provisioning.

"Rackspace Spot made it incredibly easy to spin up resilient clusters in minutes—whether through Terraform, UI, or API. It just worked, and that allowed us to focus on building, not managing infra. Their transparency around pricing and availability gave us the confidence to go all-in on spot instances. That level of visibility is rare—and invaluable."

OpsMx Engineering/ DevOps team

Business Outcomes: Modernized Infra—Lower costs, faster provisioning, more Flexibility

Having used Spot for more than 18 months to date, OpsMx has realized significant business value, including

  • Cost reduction: Achieved ~83% reduction in infrastructure costs for test/demo workloads. This optimization allowed for more efficient resource allocation.
  • Improved speed: Cluster provisioning time was dramatically improved from hours to minutes, accelerating development and testing cycles.
  • Streamlined management: Infrastructure management was streamlined with built-in monitoring and auto-healing capabilities, reducing manual intervention.
  • Scalability & flexibility: The solution enabled scalable usage without vendor lock-in concerns, providing OpsMx with long-term agility.

In addition, OpsMx has seen the product evolve rapidly, overcoming some early challenges in reliability and feature availability.  For e.g. feedback from the OpsMx team was instrumental in Rackspace Spot’s evolution to provide:

  1. A simple, easy to use Terraform provider.
  2. Improved cluster provisioning performance and reliability.
  3. Improved cluster lifecycle management.

"With the Gen-2 Kubernetes control planes, Rackspace Spot has closed the gap with top-tier cloud providers. We now get the reliability and performance we need, without the overhead or high costs—and that's been a huge win for our team."

— Sumeet Kulkarni, Engineering manager, OpsMx.

Looking ahead

Given the success OpsMx has achieved to date, they are looking to increase Spot Instance usage in production environments.  In addition, they are also interested in deeper integration with Rackspace Spot for automated Kubernetes cluster governance and security.