Portworx sponsored The New Stack’s coverage of KubeCon+CloudNativeCon North America 2019.
Last week at KubeCon+CloudNativeCon North America 2019, enterprise Kubernetes management platform provider Kublr launched what it claims is the first multi-site Kubernetes orchestration platform. With Kublr 2.0, users will be able to manage distributed clusters across multiple data centers, regions, and clouds, enabling high availability and failover. The company also unveiled two additional offerings — Kublr Accelerator and Kublr Cloud — which together help organizations to jumpstart their DevOps and Kubernetes journey.
“Kublr embeds governance, security and scalability ensuring clusters deployed by Kublr are reliable, survive control plane failures and disconnections, [and are] self-healing in all environments that Kublr supports,” said Kublr Chief Technology Officer Oleg Chunikhin.
Kublr works across Amazon Web Services, Azure, Google Cloud Platform, VMware vSphere and vCloud Director, and bare metal.
Kubernetes already offers a multisite management solution, but those clusters aren’t connected to each other. There’s a lot of DIY using open-sourced tools as organizations struggle to create some sort of inter-cluster connectivity, application, ingress traffic management, and data replication.
Kublr 2.0 provides out-of-the-box multisite orchestration with a centralized control plane to manage all your clusters across different clouds and regions. It helps connect those clusters on different levels — from network connectivity to discoverability — so those clusters and services can find each other by name and incoming traffic can be automatically routed.
“Data aggregation across all sites before was a custom project for companies. [Kublr 2.0] helps set up data storage systems that provide data replication across different sites,” Chunikhin said.
Not only does Kublr integrate with other solutions, but it also provides a way to further prevent vendor lock-in. All interactions on top of Kublr are directly with Kubernetes. To migrate your applications to a different Kubernetes provider, you simply spin up a new Kubernetes cluster, change the address in your CI/CD pipeline, and redeploy your apps in the new cluster — provided the new platform supports upstream compliant Kubernetes.
Kublr 2.0 is currently in the private technical preview stage and set to release in the first quarter of 2020. Kublr Enterprise Platform subscriptions will now be available with a new Source-Available License. Organizations can leverage this source code as a foundation for internal development, while still getting support from the Kublr team with Kubernetes deployment and operations.
“Multidata center, multicloud or multiregion will be the next essential requirement, once companies run reliable Kubernetes across several locations and data centers,” said Kublr CEO Slava Koltovich. ”To date, there’s no solution on the market that provides reliable orchestration of Kubernetes clusters across different locations.”
Kublr has already seen customers implementing this orchestration across multiple regions in multimillion-dollar projects.
Talking about both Kublr Accelerator and Kublr Cloud, Koltovich said, “We see stark differences as to how quickly enterprises can adopt a cloud native architecture. And not everyone is ready to go all-in by themselves. First, applications must be migrated, and DevOps processes set up on top of Kubernetes.”
These new tools act as a sort of Kubernetes starter kit.
Kublr Accelerator is a set of packaged services to build on top of a Kubernetes and cloud-native stack. It includes tools for:
- Application modernization and containerization
- Migrating monolithic apps to microservices
- DevOps processes and CI/CD setup
- App monitoring, failover, and scalability
Kublr Cloud is a fully managed offering of the existing Kublr Platform. Its goal is to help organizations to get up and running with both Kublr and Kubernetes. Customers starting with a managed offering can later transition to self-managed if desired at their own pace. KubeCon, Kublr Cloud is a managed service on top of the customer’s infrastructure, made to obscure the inherent complexity of Kubernetes orchestration. Kublr Cloud comes with a guaranteed SLA, including 24/7 cluster monitoring, issue identification, and resolution.
In addition, with the goal of guiding customers’ transition from monolithic architecture, Kublr offers workshops and other education events with the hope that after a year or so, organizations can run it themselves using Kublr Cloud managed service.
KubeCon+CloudNativeCon is a sponsor of The New Stack.