How DigiCert Secured 99.99% Reliability and Global Scale | Harness Case Study

Executive Summary

When you secure the digital backbone of the Fortune 100, 'trust' isn't just a marketing slogan - it's the heart of your product. But here’s the problem: how do you move fast enough to stay competitive without breaking the very trust your customers rely on?

The challenge of maintaining this level of trust was compounded by the need to scale globally. To meet the demands of their customers, DigiCert committed to moving from a 99.9% ("three nines") to a 99.99% ("four nines") uptime Service Level Agreement (SLA). Achieving this while accelerating velocity required drastically reducing the risk of change - eliminating the risks associated with manual deployments and fragmented tooling.

By partnering with Harness, DigiCert transitioned from disjointed tools to a unified software delivery foundation. Leveraging Harness Continuous Delivery, Feature Flags, and now Infrastructure as Code Management (IaCM), DigiCert successfully shifted their engineering culture from "firefighting" to "innovating," reducing the time to stand up new global regions from weeks to minutes.

"There are companies the world knows, and then there are the companies the world depends on. When your mission is securing the world, your foundation must be unshakable." — Wade Choules, SVP of Engineering, DigiCert

The Challenge: When 99.9% Isn't Good Enough

For years, DigiCert operated with a standard "three nines" of reliability. But as their customers’ needs evolved (powering airports, hospitals, and financial systems), the cost of downtime became incalculable.

"Three-nines uptime wasn’t good enough anymore, not for the customers who trusted us," says Wade Choules. "We took on the challenge of moving to a four-nines SLA."

However, their existing software delivery processes were a bottleneck to this goal:

• High-Risk Deployments: Nobody gets into engineering to stare at a progress bar or manage a crisis call at 2 AM. Yet, that was the reality at DigiCert. Deployments were high-stress events filled with cross-team handoffs and late-night firefighting. They were spending too much energy keeping the lights on and not enough time innovating.
• Fragmented Tooling: Different teams used custom scripts for different environments (Kubernetes, AWS, On-Premises), making standardization impossible.
• Global Delays: Expanding into new geographies, such as setting up a new data center in new regions, was a slow and manual process that required weeks of configuration.

DigiCert needed more than a tool; they needed a way to make reliability reproducible at a global scale.

The Solution: A Foundation for Trust

DigiCert selected the Harness Platform to serve as the unified "operating system" for its software delivery. 

"We weren’t looking for a tool. We were looking for a foundation. Something that could support every team, every environment, every region we were moving into." — Jared Daniels, Principal Software Engineer, DigiCert

Carefully rolling out changes, progressive delivery, is core to the strategy. Progressive deployment techniques ensure that changes are rolled out to only some customer traffic at a time, and impacts are measured before a full rollout. Rapid rollbacks are also a feature of progressive deployment. DigiCert uses Harness Continuous Delivery to achieve progressive delivery at the service level, and Harness Feature Management and Experimentation at the feature level.

1. Unified Continuous Delivery (CD)

Harness replaced a web of custom scripts with a single, governed pipeline technology that works consistently across DigiCert’s hybrid infrastructure.  Harness enables scriptless progressive delivery with blue/green or canary deployments that natively manage traffic shifting. Whether deploying to an on-prem data center or a Kubernetes cluster in AWS, the developer experience remains the same. This consistency enabled DigiCert to transfer deployment ownership from a central Ops team to engineering, thereby eliminating bottlenecks and time zone delays.

2. De-Risking with Feature Flags

To achieve four-nines reliability, DigiCert needed to decouple deployment (moving code) from release (exposing features). Using Harness Feature Management and Experimentation, engineering teams can now deploy code safely behind a toggle, test it in production with a small subset of users (Canary deployment), and roll it back instantly if an anomaly is detected, without redeploying the binary.

"It let us shift from firefighting to innovating," says Kelsey Eiman, Director of Software Engineering. "Deployment-time incidents, which used to keep us up at night, became rare."

The Next Frontier: Scaling Infrastructure (IaCM)

With software delivery velocity solved, DigiCert faced a new bottleneck: the infrastructure itself. While they could deploy code in minutes, provisioning the underlying infrastructure for a new global region was still a manual, multi-week process.

To close this gap, DigiCert is adopting Harness Infrastructure as Code Management (IaCM).

By bringing the same pipeline rigor to infrastructure as they did to software, DigiCert creates a "cookie-cutter" model for global expansion. Instead of manually configuring servers and networks for a new region, they can treat their data centers as code—versioned, governed, and automated.

The IaCM Impact: Weeks to Minutes

"We haven't stopped just with continuous deployment," explains Wade Choules. "We are now leveraging Harness Infrastructure as Code to spin up our data centers around the world. Prior to using Harness, standing up a new region would have taken us weeks. With Harness, we did it in minutes."

This capability creates a virtuous cycle: Harness IaCM provisions the "hardware" instantly, and Harness CD deploys the services immediately after, allowing DigiCert to "copy-paste" their presence to any region the world demands.

The Results: From Firefighting to Innovating

The transformation at DigiCert has turned compliance and reliability from a tax into a competitive advantage.

• Four Nines Achieved: The shift to automated, governed pipelines was a key enabler in meeting their new 99.99% uptime SLA.
• Global Agility: The time required to stand up services in new global regions dropped from weeks to minutes.
• Engineering Velocity: By removing the "toil" of manual deployments and tickets, developers reclaimed hours of productivity every week.
• Empowered Teams: Engineering teams now own their own pipelines. 
• Consistent Trust: A single, unified process for On-Prem, AWS, and Kubernetes ensures that security standards are enforced automatically, regardless of the environment.

"Harness didn’t just improve our pipelines—it helped us achieve four-nines uptime and operate with the level of predictability we need at a global scale. Harness helps us deliver trust, every deployment, every region, every day." — Wade Choules, SVP of Engineering, DigiCert