MakerBot believes there’s an innovator in everyone. As a global leader in 3D printing, they set the standard in reliability and ease-of-use by providing effective solutions for every stage of the 3D printing process. From the start, they have redefined the possibilities with 3D printing for users across all industries.
Open Source Continuous Delivery Dilemma
MakerBot’s software delivery process kept Erik Ahrend up at night, literally. As the Lead Cloud Architect, Erik maintained MakerBot’s Spinnaker pipelines. Erik spent three hours a day troubleshooting deployment issues and often was woken up in the middle of the night to fix broken deployments. The total cost of Erik’s time added up to roughly $118,000 a year.
Spinnaker pipelines and API’s required advanced knowledge to create and edit. Erik was the only one with enough experience to fix deployment issues. MakerBot hired Armory to help manage their CI/CD pipelines. Unfortunately, MakerBot’s deployment issues persisted.
Deployments had an unexplainable 48-hour lag between uploading an image and reaching production. Most developers found it easier to manually update Helm than wait for Spinnaker. Erik needed to fix and simplify MakerBot’s software delivery process.
The Cost of Custom Governance
In addition to simplifying deployments, Erik needed to add capabilities. The pipelines lacked basic integrations with APM tools like DataDog, which slowed down the verification process. The pipelines also lacked governance. Without RBAC controls or audit trails, Eric feared for the safety of MakerBot’s applications.
Erik planned to spend 6 months coding these two features and adding them to MakerBot’s pipelines. The project would require at least two developers and cost MakerBot $150,000 in developer effort.
Realizing the exorbitant costs in front of him, Erik decided to search for a company that could provide out of the box solutions.
Harness Provides Priceless Confidence
Erik turned to Harness for that solution.
Harness created a self-service deployment culture. Developers are able to copy Harness’s pipeline as code into whatever new pipeline they need, and Erik doesn’t have to spend 3 hours a day babysitting deployments. Harness also provided MakerBot with advanced RBAC controls, Datadog and Jira integrations, and templatized canary deployments. These capabilities reduced MakerBot’s security concerns and increased deployment confidence.
In just 4 months using Harness, Makerbot is deploying ~15 times per day with a very low failure rate of only 8%. Most importantly, Harness was able to provide Erik with some much-needed headspace.