CI/CD Highlights from KubeCon Europe 2021
KubeCon+CloudNativeCon Europe 2021 – Virtual wrapped a couple of weeks ago, and we now have a treasure trove of conference talks to watch at our leisure. There are 225 of them!
I’ve managed to catch up with some of these. While there are always quite a few takeaways from an event of this size, I’d like to focus on one fairly obvious one: The focus on continuous integration/continuous delivery (CI/CD) continues to increase as enterprises prioritize their DevOps approach.
Here, I’d like to highlight three talks that I think are worth watching from the recent KubeCon. I will then draw some conclusions that I think may be of value for organizations evaluating their CI/CD needs and approaches to building pipelines.
Kara de la Marck and Fatih Degirmenci highlight the numerous benefits of interoperability in CI/CD and how it can assist in managing complexity and making changes easier. The Continuous Delivery Foundation is doing some excellent work in making this landscape more standardized and approachable.
In the above talk, Juergen Etzelstorfer and Karthik Satchitanand dig into the importance of parity between dev and production environments and how LitmusChaos can provide much-needed help with overall application resiliency. They advise: “Break all the things in development, not production.”
Vibhav Bodade and Vincent Demeester walk through various aspects of debugging Tekton. Related to this, I’d like to point out that IBM has a useful tutorial on creating a Tekton-based delivery pipeline to deploy a simple app to Cloud Foundry.
Takeaways
There is a lot happening in this space right now and for very good reason. Continuous delivery improves productivity and deployment velocity, and decisions made about your CI/CD approach have tremendous implications for your organization’s overall effectiveness. This is an increasingly dynamic and complicated space that no modern tech team can ignore.
Selecting which cloud native approaches, including CI/CD tools, are right for your organization is inherently quite complex, even daunting. In some cases, this complexity allows for valuable customizations that will fit with your specific business needs. However, in other cases this complexity might be an unnecessary albatross, slowing development and hindering essential work. In the latter case, an opinionated platform, such as Cloud Foundry, might best fit your needs.
Blue-green deployments, canary nodes and out-of-the-box interoperability with your favorite CI/CD tools are some of the essential baselines I’d recommend for any enterprise assessing their DevOps needs. Tekton, Circle CI, Travis CI, GitHub Actions, Concourse CI — all either already have native support for the Cloud Foundry API or can be integrated quite easily.