Ubiquitous Cloud-Native Service (UCS)

As containerized cloud-native applications become deeply integrated into various business scenarios within enterprises, new demands are increasingly prominent. These demands include cross-cloud and cross-region unified collaborative governance, as well as ensuring consistent application experiences. To meet these demands, Ubiquitous Cloud-Native Service (UCS) has emerged, thereby accelerating the cloud-native upgrade across various industries.

UCS is a platform for the centralized management of distributed clusters and enables unified collaborative governance of containerized cloud-native applications across clouds and regions. It runs on Karmada, the CNCF’s first multi-cloud container orchestration project, and supports unified management of T Cloud Public's cloud infrastructure (CCE clusters, CCE Turbo clusters), on-premises infrastructure (self-managed Kubernetes clusters), and third-party cloud infrastructure (other cloud vendors' Kubernetes clusters).  

UCS provides enterprises with a consistent experience across all domains, encompassing the deployment and management of cloud-native services as well as the application ecosystem. This comprehensive coverage spans a wide range of use cases, including core zones, hotspots, customer data centers and business sites. This ensures that customers are not constrained by geographical location, cross-cloud limitations or traffic restrictions when using cloud-native applications. By bringing cloud-native capabilities to every business scenario within an enterprise, UCS accelerates the adoption of cloud-native technologies across various industries.

UCS supports unified access and management for clusters across clouds and regions. It can connect to T Cloud clusters and attached clusters.

Based on the Karmada kernel, UCS can achieve unified access for thousands of distributed clusters, enabling collaborative scheduling of millions of node resources. With response times measured in seconds, UCS can optimize application deployment based on global resource distribution and business characteristics, considering factors such as geographical location, Quality of Service (QoS) and resource balancing.

UCS deeply integrates the capabilities of T Cloud Public's Image Repository for Container (SWR). It supports full lifecycle management of images, providing users with simple, easy-to-use, secure, and reliable image management functions to help them quickly deploy containerized services.

UCS Permission Management is built on the capabilities of T Cloud Public's Identity & Access Management Service (IAM) and Kubernetes' Role-Based Access Control (RBAC). It not only supports the fine-grained permission control based on IAM and IAM Token authentication, but also supports permission control at the cluster level, container fleet level, cluster namespace level, and container fleet namespace level.

UCS supports unified access and management of Kubernetes clusters across regions and clouds (T Cloud Public, On-Premises, and third-party clouds) scenarios. Additionally, it provides permission management for enterprise-level project tenants, enabling fine-grained management of sub-account access permissions to Kubernetes resources. This facilitates compliance audits for multi-cloud and multi-cluster environments.

UCS is developed based on the multi-cluster management project Karmada. Through the capabilities of a multi-cloud management platform, it can support the unified access of thousands of distributed Kubernetes clusters.  

Furthermore, it can support the collaborative scheduling of millions of node resources, enabling elastic scaling of applications across clouds and clusters in distributed cloud scenarios. It also facilitates cross-cloud migration of applications in the event of a failure. Based on the global distribution of resources and the characteristics of the business, it can optimize the deployment of applications according to conditions such as geographical location, network QoS, and resource balance. This allows users to provide the computing power resources required by applications at any time and place.

With the unified traffic entry point MCI (Multi Cluster Ingress), the boundaries between regions and clusters have been broken down, an intelligent scheduling hub spanning clouds and regions has been established for enterprises. 

It not only significantly reduces the complexity of managing multiple clusters but also achieves high availability and elastic scalability of services through refined traffic routing policies and automated health checks. Combined with full-stack TLS security protection, it ensures flexible and efficient traffic distribution while providing continuous assurance for globally distributed applications.