Learn how Kubiya is Developing Temporal Services Faster with Velocity
Kubiya, a leading generative AI for DevOps startup with a customer list that includes the likes of AWS, Blu and Aston Martin to name a few, has embraced Velocity as a daily dev tool to dramatically speed up their development cycles.
Central to Kubiya's workload is Kubernetes development, but it had become a bottleneck. For example, in an effort to enhance error handling and retries, Kubiya's engineers integrated Temporal, a super powerful distributed open-source platform, into their microservices via an SDK.
“We're using Temporal for specific error handling and retry abilities,” shared Noam Raz, a lead engineer at Kubiya. While this integration has significantly bolstered their application's robustness, development within Kubernetes became sluggish, hindering new feature implementation and bug fixes alike.
Each Kubiya developer had their isolated dev environment containing microservices, secrets, and configurations. Committing changes in Git triggered CI/CD processes, which ultimately deployed the new code to the dev environment, but this traditional, industry-standard approach to developing for Kubernetes is painstakingly slow, leading to constant context switching for developers, and extended dev cycles.
This development process was very slow. - Noam Raz
And while the intermediate solution of using kubectl port-forward made it possible to develop Kubiya's remote services locally, it was ultimately too brittle and unreliable a solution for extended development sessions.
Enter Velocity, an IDE-first development tool, that revolutionized Kubiya's approach. Velocity empowers developers to work locally in their IDE, with code changes continually syncing with a deployed service in Kubernetes. This eliminated the need for continuous CI/CD workflows during local development and testing.
Once I configured Velocity, the execution was extremely simple. - Noam Raz
Developers could now develop, debug, and test code running in Kubernetes directly from their IDE. And they could commit changes only after ensuring the code worked as expected.
The deploy process was much faster, and Velocity even enabled me to debug remotely, which is a great feature. - Noam Raz
Kubiya's development cycles drastically shortened with Velocity becoming an integral daily tool for the team.
It became one of my daily dev tools. - Noam Raz
Velocity's ability to sync a local debugger with Kubernetes code proved invaluable. “The ease of developing and investigating problems in production was remarkable,” noted Noam.
Moreover, Velocity empowered developers to self-serve, and develop on top of DevOps existing configurations and infrastructure, extending time savings further without extra cognitive load.