CNCF Provides Access to Reference Architectures for Cloud-Native Use Cases
The Cloud Native Computing Foundation (CNCF) has published a report that defines guidance reference architectures, in addition to guidance to help IT teams better identify which open-source software projects are mature enough to be used for specific use cases.
Defined by the End User Technical Advisory Board, the CNCF Technology Landscape Radar report initially identifies technologies that have been proven ready to be used to enable application management across multiple clusters along with which tools and platforms lend themselves best for batch-oriented/machine learning/artificial intelligence (AI) applications.
Taylor Dolezal, head of ecosystem at CNCF, said the end-user members of the cloud-native community have long sought more examples of how various open-source technologies in combination with one another can be reliably applied to drive specific applications. The End User Technical Advisory Board is now taking responsibility for that effort in a vendor-neutral forum, he added.
The initial reference architectures being made available are Scaling through Platform Engineering at Allianz Direct and Scaling Adobe’s Service Delivery Foundation with a Cell-Based architecture. In addition to detailing how these architectures are being used in production settings, they include insights into the context an organization used to define them. Six more reference architectures are in the process of being approved, noted Dolezal.
The CNCF End User Community is made up of practitioners from more than 150 organizations that meet regularly to share adoption best practices and provide feedback on project roadmaps and any future project the CNCF might sponsor. Members of the advisory board are either nominated by CNCF sponsors or elected by the members of the community. The CNCF End User Community partnered with SlashData to survey 340 professional cloud-native application developers to conduct a survey to identify the most mature multicluster application management and batch/AI/ML technologies.
Cilium, a networking platform based on extended Berkley Packet Filtering (eBPF), received the highest maturity score (+47) for multicluster applications, while Apache Airflow (+51) had the highest maturity score for AI/ML technologies.
It’s not clear how applicable each of these ratings is to specific organizations, however, in an era where a project ranked as being at the sandbox level by the CNCF is actually a mature technology, the guidance provided by the CNCF End User Community adds some much-needed context. Otherwise, organizations might decide to limit the adoption of some technologies to only graduated technologies, when other projects may have already proven their mettle in production environments. There are, after all, more than 200 CNCF projects spanning everything from foundational technologies such as Kubernetes to a host of add-ons that not every IT team is going to necessarily require.
Hopefully, the collective experience of the CNCF End User Community will lead to additional enhancements being made to projects that will increase the pace at which they become mature enough to use in production environments. The one certain thing is that experiences being shared by the CNCF End User Community are most definitely only the kind that can be hard-won.
