IBM Corp. today pitched a new open-source project intended to help architects, developers and operations teams collaborate more closely on building Kubernetes-based software applications.

The project, introduced at the O’Reilly Open Source Software Conference in Portland, Oregon, is called Kabanero. It comprises a number of new open-source tools, including Appsody, Codewind and Razee, combined with existing tools such as Knative, Istio and Tekton. The combined tools provide an “end-to-end solution for you to architect, build, deploy, and manage the lifecycle of Kubernetes-based applications,” IBM said.

Kabanero is all about lowering the barrier to entry for companies wanting to adopt Kubernetes software that’s used to orchestrate large clusters of containers, which in turn are used to host modern applications.

“Choosing the right technology for building cloud-native apps and gaining the knowledge you need to effectively adopt Kubernetes is difficult,” Nate Ziemann, senior product manager for IBM Cloud’s Developer Technologies, wrote in a blog post. “On top of that, enabling architects, developers and operations to work together easily, while having their individual requirements met, is an additional challenge when moving to cloud.”

One of the biggest challenges enterprises face is that architects and operations teams must ensure that things such as their security standards are incorporated into their Kubernetes applications. This requires a great deal of cooperation between those teams and their developers, but of course doing so can also slow down the development pipeline. To address this, Kabanero incorporates a command line tool called Appsody, which can be used by developers to create microservices, the components of containerized apps, which already meet their organization’s security standards.

“Appsody gives you pre-configured stacks and templates for a growing set of popular open source runtimes and frameworks, providing a foundation on which to build applications for Kubernetes and Knative deployments,” Ziemann said. “You can customize Appsody stacks to meet your specific development requirements and to control and configure the included technologies. If you customize a stack, you have a single point of control from which you can roll out those changes to all applications built from them.”

Another key component of Kabanero is Codewind, which provides extensions to popular integrated development environments such as Eclipse, Eclipse Che and VS Code, which are used by developers to build their apps. The idea with Codewind is to reduce the learning curve for developers, enabling them to get started on building Kubernetes apps in an environment they’re already familiar with.

Kabanero also makes use of another open-source tool called Razee, which provides “multi-cluster continuous delivery tooling for Kubernetes” and is used to manage the progression of apps from development and testing all the way to production deployment.

“Kubernetes has won out as the leading container platform, so now companies are more focused on helping developers build their next-generation applications efficiently,” said Holger Mueller, principal analyst and vice president of Constellation Research Inc. “IBM’s game plan is to attract developers through tooling and derive workloads for its execution architectures from that. It will be interesting to see how far adoption of these tools goes in a few months.”

Open-source datasets

In addition to Kabanero, IBM also announced the launch of a new Data Asset eXchange for data scientists and developers who need access to open datasets in order to train artificial intelligence and machine learning algorithms.

DAX, as it’s called, is a collection of “carefully curated free and open datasets” that come with “clearly defined open data licenses,” IBM said.

Similar resources already exist on service such as GitHub, but IBM said DAX is unique because all of the datasets and their metadata are available in a standardized format, which means they’re more straightforward to adopt.

Analyst Doug Henschen, also of Constellation Research, told SiliconANGLE the datasets offered by IBM would still need to be vetted by developers to ensure they’re relevant to their specific environments. But he also had praise for the initiative.

“It’s nice to hear that these data sources are curated, open, standardized in terms of formats and metadata, and industry specific, so they may provide value, particularly in the early development phase of applications,” Henschen said. “As developers move toward refining models that are specific to their businesses and applications, that’s where data at scale and data specific to an organization becomes invaluable. It’s also where Constellation is seeing a lot of activity around the development of synthetic data and use of adversarial networks seeded by organizations’ specific data.”

Image: Kreatikar/Pixabay