Market Problem

OpenShift customers require a growing list of advanced Kubernetes networking capabilities for their business-critical production deployments. This captures cross-industry and customer-driven prioritizations of those OpenShift Advanced Networking requirements that may include, but are not limited to, telco customers.

Advanced Networking use cases were certainly catalyzed by adoption and growth into the Telecommunications market segment – advanced networking is their business, after all – but requirements have also come from layered products (e.g. OpenShift Virtualization and Service Mesh) whose customers have come to expect sophisticated networking configuration, portability, hardware integration, performance, protection and observability. OpenShift's traditional market segments, including enterprise/IT customers, are now also directly benefiting from these advancements, and they are realizing more applications and reasons to continue to grow their deployments on OpenShift, as a result.

Why it Matters

OpenShift requires a flexible open source networking solution with community involvement and with solid underpinnings to grow upon and scale to our current and future customer use cases, across all market segments.

A major strength of OpenShift is the breadth of platform integration. Customers require "run anywhere" but they also require platform-native capabilities and integration.

Illustrative User Stories or Scenarios

Choosing a single illustrative example to demonstrate a common networking complexity entanglement: load balancing. While OpenShift does not provide enterprise-grade load balancing, we certainly must integrate with and provide best practice support for a default, best-practices load balancer solution for every platform we have an integration for. Native load balancers are generally unique across public and private cloud providers, and a single cloud provider may have many load balancers to choose from (e.g. ELB, NLB, ALB on AWS), or there may be a reliance on a 3rd-party solution (e.g. F5) for a performant/enterprise-grade solution on some platforms. OpenShift cluster load balancers are "not an island, entire of itself", customer deployments are global and span multiple clusters, and so our scope for configuration and functionality must match that global perspective.

Common Advanced Networking infrastructure use-case complexities this Market Problem is focused upon, both locally and globally (centralized/unified management):

• security (modern encryption tooling, separation of sensitive internal-only traffic from data plane or pubic traffic, auditing for regulatory compliance)
• multi-cluster
• hybrid cluster
• edge (telco and enterprise)
• far-edge (telco)
• remote worker node deployments
• HyperShift architecture
• many-node vs single-node deployments
• VNF to CNF transformation enablement
• customer-required 3rd-party primary networking plugins
• secondary networking plugins (Multus-fueled)
• vendor hardware (NICs, SmartNICs) and offload capabilities
• high-performance networking
• unification of configuration
• observability of complex data and traffic analysis

Expected Outcomes

• An open source networking solution, "out of the box", that meets the requirements of the most demanding customer use cases from each market segment
• Extensibility to enable bring-your-own and 3rd-party networking solutions
• Multi-cluster feature scope and management
• Improved observability of the network
• Internal infrastructure to quickly and efficiently support the use cases

Effect

• Customer retention – out network and its capabilities grows ahead of customer use cases
• New markets – remove barriers to adoption for new use cases and markets
• Improved experience
  • simplicity of understanding complex network deployments,
  • multi-cluster scoped visibility, management, and workloads,
  • unification of configuration and traffic management

Partner

• Service Mesh
• OpenShift Virtualization
• 3Scale
• Telco-specific teams
• ACM
• Installer