Raisin Runs from Jenkins Nightmare – Saves 525k per Year

About

Raisin is the leading pan-European wealth management platform, connecting retail customers with financial institutions looking to expand or diversify their deposit reach. They are available in English across Europe and operate country-dedicated platforms in Germany, France, Spain, the United Kingdom, Ireland, the Netherlands, and Austria. In 2020, they launched a savings-as-a-service branch in the U.S.

Modernize, They Said… It’s Easier and Cheaper, They Said…

Raisin migrated its applications to the cloud to take advantage of modern technologies like Kubernetes.

Makes sense. Everyone is doing it.

A side effect of their application modernization was the need for a new software deployment strategy to reduce microservice deployment effort.

Issue identified.

Raisin created a central DevOps team led by José Meyer to tackle the deployments.

Plan created.

The team used Jenkins and custom scripts to build deployment pipelines for various developer teams.

Problem solv… hold up. Incoming migraine.

The Jenkins pipelines allowed Raisin to deploy 10 times every month, but the 6 new members of the DevOps team spent 80% of their time keeping Jenkins standing. That equates to roughly $700k per year in salaries (assuming $70 per hour per team member). Between Jenkins maintenance, updates, plugin additions, and troubleshooting, the team hardly had time for anything else. Jenkins was too fragile to be left on its own.

Jenkins scripts were sensitive and often caused deployment failures. - José Meyer | Lead CI/CD Engineer | Raisin

Because of the Jenkins pipeline complexity, onboarding a new application took roughly 2 days. That doesn’t sound terrible on it’s own, but figure in the massive maintenance effort after onboarding and you’re talking about a major time suck.

José and the DevOp team initially picked Jenkins because it was a “free” solution. But after dealing with the Jenkins nightmare, Jose began looking for better ways to deploy software.

Even though Jenkins is an open-source tool, the cost of maintaining it (updates, plugins, etc.) exponentially displaces any potential savings. - José Meyer | Lead CI/CD Engineer

A Modern CD Solution for a Modern Architecture

During José’s search, he came across Harness. With Harness, José and the DevOps team distributed software deployment responsibility to the application teams. The application teams now manage their own services, pipelines, and triggers. José’s team of 6 now only spends 20% of their time administering software delivery. That’s a 60% reduction in effort worth roughly $525k.

We have drawn clear lines of ownership using Harness. My team owns workflows and templates, and each application team manages their own services, pipelines and triggers. - José Meyer | Lead CI/CD Engineer | Raisin

The democratization and automation of software delivery have allowed Raisin to deploy 53 times a month – a 5x increase from their previous deployment velocity.

The only manual step left is the manual approval. And that’s by design. - José Meyer | Lead CI/CD Engineer | Raisin

The news of Harness’s deployment success spread quickly around Raisin. Application teams began lining up to have their app deployed using Harness. Luckily, it only takes José’s team 30 minutes to onboard a new application due to Harness’ templatization and reusable connectors.

Modernizing applications and CI/CD can be a challenge. Maybe it’s a challenge you’re getting ready to experience. Check out how Harness makes the transition less painful.