Red Hat and IBM Research have bolstered their hybrid-cloud focus with a new open source project designed to make it easier for organizations to migrate their legacy applications into a cloud-native environment.

The Konveyor Project is an overarching platform that includes several components targeted at various migration models. Those components, or subprojects as Red Hat refers to them as, use Red Hat’s Kubernetes Opertors deployment model and are focused on re-hosting, re-platforming, or re-factoring legacy applications.

The most basic path is re-hosting, which James Labocki, senior director of project management for cloud platforms at Red Hat, explained is targeted at organizations or IT departments that lack the ability to change or modify code. For this model, the Konveyor Project has its Crane and Forklift tools.

Crane is focused on migration applications between clusters running on different Kubernetes iterations or allowing users to migrate those clusters to a different underlying infrastructure like if they were moving between larger cloud deployments.

Forklift is focused on migrating virtual machines (VMs) to Kubernetes and can also migrate VMs into the open source KubeVirt platform to provide a unified approach for developers to support containerized workloads. KubeVirt can also manage VM-based workloads that are difficult to integrate into a containerized environment.

Re-platforming is a bit more involved and typically requires a change to the underlying technology used by an application but also minimizing code changes. This is often the case when an organization is moving its applications into Kubernetes.

For this model, Red Hat has the Move2Kube subproject that taps into past IBM work around migrating applications from the Docker Swarm container orchestrator or Cloud Foundry-based applications. Move2Kube translates existing artifacts into Kubernetes-native concepts to support that migration and allow those applications to run in the Kubernetes environment.

The most in-depth tool is the Tackle application that handles the more complex task of re-factoring applications. This involves making fundamental changes to an applications architecture and development in order to move it into a cloud-native environment.

All four of these tools can take advantage of the fifth, which is Pelorus. This tool measures application metrics for software delivery performance, lead time for change, deployment frequency, mean time to restore, and change failure rate to see if the other tools are performing their job as expected.

Labocki explained that these tools provide organizations with a hybrid-cloud approach to migrating their brownfield applications into a modern network architecture.

“We believe that there's an opportunity for those that are willing and interested to actually help customers in a greater way to actually open all of those tools so that people can benefit from them,” Labocki said. “But also it's not moving you into one cloud but it's moving into a hybrid cloud like deployment across multiple clouds.”

The timing of the launch to coincide with this week’s KubeCon + CloudNativeCon EU 2021 virtual event was not by chance. Labocki said that the Konveyor Project was in discussions with the Cloud Native Computing Foundation (CNCF) about a possible move into the open source group’s hierarchy.

“We’ve been talking to the SIG App Delivery group about getting guidance from them if we’er going to bring these into the sandbox,” Labocki said. “Whether we bring Konveyor as a whole under there or separate subprojects.”

The tools were developed through various meetings Red Hat hosted with its community partners. The meetings provided a place for developers to show what they were working on to tackle this overall challenge and required the use of non-proprietary technology and often a demonstration of the efforts.

Labocki added that the project has had participation from other CNCF projects on its meetups and also from other vendors like VMware and “competitors of ours coming and talking about their tooling and what they’re doing," as well as meetings with the IBM Research team.

“We think that if we can get enough people using these tools in the community then they're going to be requesting that vendors provide these tools, and then when that happens more people will get behind the same toolset as opposed to trying to use it as a proprietary differentiator,” Labocki explained.