Cloud Foundry Advisory Board Meeting, Apr 2021: Cloud Native Buildpacks
This month’s Cloud Foundry Community Advisory Board (CAB) meeting centered around Cloud Native Buildpacks, as well as discussed development updates around the ecosystem projects. The call was moderated by Chris Clark from the Cloud Foundry Foundation.
Initially created by Heroku in 2011, buildpacks provide framework and runtime support for applications. By identifying dependencies, buildpacks determine how to configure apps, so they can communicate with bound services.
In 2018, Pivotal and Heroku started the Cloud Native Buildpacks project, which aimed to unify the buildpack ecosystems. During the call, Emily Casey of VMware provided an overview of the current state of the project, as well as its goal moving forward.
According to Emily, Cloud Native Buildpacks is primarily a specifications project. A Platform API and a Buildpack API are defined under the specifications. Additionally, the project also includes a Lifecycle component, which understands both APIs. This very component is used by all the adopters of Cloud Native Buildpacks to bridge the gap between platforms and buildpacks.
The Buildpack API describes the contract a buildpack needs to comply with in order to participate in the system. Lifecycle then exposes an API to the platforms, enabling them to understand how buildpacks run, and how their output is transformed into an image.
“The goal of Cloud Native Buildpacks is to keep all the good stuff in v2 Buildpacks, while addressing the drawbacks. We want to bring this to a wider audience and hopefully make it an industry standard, so it’s not just a piece of the Cloud Foundry platform, but is a standard for builds across the industry.”
—Emily Casey, VMware
Eric Malm from VMware listed the following updates:
- The Release Integration team delivered cf-deployment v16.11.0. The team also released cf-for-k8s v3.0.0, which adds new certifacte properties related to injecting the instance index into application pods.
- The KubeCF team is reintroducing the cf-deployment-updater job. They are also improving the operation of log-cache within the memory limit.
- The CAPI team is working on app manifests being set and applied on the server side.
- The Eirini team is filling in gaps between its custom resource definition (CRD) and REST APIs. They are also looking into running apps as deployments instead of StatefulSets.
- The Networking team is modifying the Gorouter to check only Subject Alternative Names (SANs) for hostnames. This resolves some TCP router and policy server issues.
- The HAproxy team released fixes for Bionic stemcell compatibility issues.
Eric also brought up an ongoing discussion around the vision for Cloud Foundry on Kubernetes and asked for community feedback on the current proposal. Anyone interested in participating in the next live discussion can join the upcoming SIG meeting on May 4, 2021, at 8:30 a.m. PDT.
Chris noted that there are already over 200 attendees registered for the next virtual Cloud Foundry summit, which will take place on July 21–22, 2021. The call for proposals closes on May 21, while voting for track cochairs ends on April 25.
He also listed some notable releases in his latest biweekly technical roundup. Chris highlighted the confirmation of two new project leads: Amelia Downs and Guiseppe Capizzi of VMware for the Networking and Eirini teams, respectively.
During the call, he urged community members to check out a recent presentation led by CFF Developer Advocates Shedrack Akintayo and Ram Iyengar, as well as VMware Product Manager, Kashyap Vedurmudi. The trio demonstrated how to improve cloud-native continuous integration and delivery pipelines with Tekton and Paketo Buildpacks.
Want more? Watch the video!
In this video, the Runtime PMC teams provide updates around ongoing projects as of April 13, 2021.