Cloud-native and Kubernetes solutions
An application delivery and security platform can unify traffic management, security, and observability for Kubernetes, helping organizations deliver and scale apps, APIs, and AI across environments with less cost, complexity, and risk.
Move fast without losing control
The F5 Application Delivery and Security Platform (ADSP) gives developers agility while unifying Kubernetes traffic management, security, and observability to deliver and scale apps and AI with consistent policies everywhere.
Standardize and secure AI delivery everywhere
Accelerate developer agility in Kubernetes and cloud-native environments for AI in production, standardizing app, API, and inference traffic across any environment
Predictable performance from data ingestion to inference
Optimize GPU cluster utilization and accelerate AI initiatives by eliminating data pipeline and traffic bottlenecks in production
Unify governance across all environments
Centralizing policy, security, and observability reduces cost and provides end-to-end governance across clusters, tenants, teams, core, cloud, and edge
Safely accelerate service monetization
Minimize business risk with processes and safeguards designed for carrier-grade multi-tenant networks and AI
Understanding CNF and Kubernetes networking, inference, and security use cases
Networking for Kubernetes
Networking for Kubernetes
Kubernetes solves core operational challenges, but it does not control how traffic behaves. As organizations manage multiple clusters, tenants, and teams, this control can lead to fragmented tooling, inconsistent security, limited end-to-end visibility, and systems that are harder to operate. The F5 Application Delivery and Security Platform (ADSP) extends enterprise-grade solutions for networking, security, and observability into modern Kubernetes environments. With a centralized control layer, teams can standardize Kubernetes networking, support AI inference workloads, and scale securely without operational overhead.
Learn more
F5 Gateway Fabric for NGINX Enterprise
A lightweight reverse proxy, load balancer, and API gateway deeply embedded in cloud-native architectures for delivering lightweight microservices in Kubernetes
F5 BIG-IP Next for Kubernetes
Delivers advanced capabilities across deterministic control of ingress and egress, microsegmentations, and strict multi-tenancy, and high-performance traffic handling for large-scale or latency-sensitive workloads
F5 Container Ingress Services for BIG-IP (CIS)
Connects Kubernetes resources to BIG-IP to manage ingress traffic, enforce consistent application delivery and security policies, and support containerized applications using familiar BIG-IP operational models
Application security for Kubernetes
Application security for Kubernetes
F5 delivers unified, container-native protection for web applications, APIs, and AI workloads directly inside Kubernetes, enforcing consistent security at ingress and between services. F5 blocks common web attacks, bots, and API abuse, secures access with identity and encryption, and applies rate limits to prevent overload. Centralized policies and visibility make protections consistent across clusters and environments, helping teams ship faster while reducing risk and operational complexity.
Watch a demo
F5 Container Ingress Services for BIG-IP (CIS)
Automates, synchronizes, and centralizes policies, WAF protections, TLS, and visibility across services and clusters
F5 WAF for NGINX Enterprise
Delivers native, lightweight, and high-performance Layer 7 protection, embedding WAF protections at Kubernetes ingress, blocking OWASP threats and bots while supporting DevOps automation and consistent policies
F5 BIG-IP Next for Kubernetes
Delivers high-throughput, scalable ingress security with ingress/egress traffic management, TLS, policy enforcement, and multi-tenant isolation for Kubernetes apps
AI inference on Kubernetes
AI inference on Kubernetes
F5 ADSP simplifies AI inference on Kubernetes, optimizing network traffic flows by steering requests to the best available inference endpoints, dynamically balancing load across clusters and GPUs, improving performance, minimizing latency, and ensuring high model availability. It secures access with consistent identity, policy enforcement, encryption, and threat protections for APIs and model services. The platform provides robust security guardrails to defend endpoints against emerging exploits and unauthorized access, enabling enterprises to scale and run their modern AI workloads with total confidence. F5 ADSP also optimizes inference traffic with smart routing, connection reuse, rate controls, and end‑to‑end observability, improving performance, minimizing latency, driving reliability, and ensuring high model availability and superior GPU utilization as demand scales.
Learn more about AI factory load balancing
F5 Gateway Fabric for NGINX Enterprise
Streamlines AI inference through dynamic Kubernetes gateway routing, load balancing, and policy controls, steering inference traffic reliably across GPU-backed services and clusters while rate-limiting high-volume LLM API traffic
F5 BIG-IP Next for Kubernetes
F5’s containerized gateway accelerates, load balances, and secures Kubernetes inference services, optimizing L4–L7 traffic, TLS, and policies for reliable GPU utilization
Cloud-native functions (CNFs)
Cloud-native functions (CNFs)
Cloud-native functions help enterprises and service providers deliver advanced networking and security capabilities quickly in a cloud-native form factor, improving resilience while reducing cost and operational friction. The F5 Application Delivery and Security Platform (ADSP) enables this by unifying traffic management, security, and observability across Kubernetes and hybrid environments. Enterprises benefit from DevOps-driven updates, consistent hybrid and multicloud policy, elastic scaling, higher reliability, and fewer point tools. Service providers gain CNF-ready services, faster API-driven rollout and monetization, simpler operations, rolling maintenance, and efficient scale to improve cost per Gbps and accelerate capacity growth.
Learn more
F5 BIG-IP Next for Kubernetes
Centralizes Kubernetes ingress / egress networking and security, unifying multi-network routing, 5G protocols, zero trust security, and high-performance AI infrastructure scaling, optimizing AI traffic, isolation, and visibility
F5 BIG-IP Cloud-Native Networks Functions (CNFs)
Cloud-native network functions deliver containerized traffic management, CGNAT, DNS, and zero trust security to dynamically scale and automate Kubernetes infrastructure, and consolidate critical services for enterprises and providers
F5 BIG-IP eBPF Observability
Real-time, kernel-level eBPF telemetry gives deep Kubernetes traffic visibility across environments, speeding troubleshooting and security insight without sidecars
Technology alliances
Resources
Articles and eBooks
Recent news
Videos
Frequently asked questions
Kubernetes is an open-source container scheduler and orchestration platform that automates the deployment, scaling, networking, and lifecycle management of containerized applications. It provides abstractions such as pods, services, and ingress to decouple applications from underlying infrastructure, enabling portability across on-prem, cloud, and edge environments. Kubernetes is widely used to run microservices, APIs, and increasingly AI/ML workloads due to its scalability, resilience, and ecosystem support. The Kubernetes project is hosted by the Cloud Native Computing Foundation (CNCF), a subsidiary of the non-profit Linux Foundation. In addition, there are many different software and SaaS vendor distributions of Kubernetes available.
Cost-effective Kubernetes starts with controlling operational overhead. A primary operational focus should be on standardizing networking, traffic management, and security rather than layering point tools per cluster or cloud. Organizations reduce operational cost by reusing a consistent ingress and Gateway API architecture, templating deployments, and improving utilization through intelligent load balancing and observability. Performance is preserved by using production-grade networking that supports high throughput, low latency, and automation via Kubernetes-native APIs and GitOps workflows.
F5 optimizes Kubernetes networking for AI/ML by providing intelligent traffic management that goes beyond basic round-robin load balancing. F5 supports model-aware routing, high-throughput L3–L7 traffic handling, and deterministic performance required for GPU-backed inference and training workloads.
As organizations expand their use of Kubernetes, it can be easy to create cluster sprawl and Kubernetes cloud service sprawl. This typically creates fragmentation across networking, security, and observability. Different ingress controllers, cloud load balancers, and policy models lead to configuration drift, inconsistent security enforcement, and limited end-to-end visibility. Operational complexity increases as teams manage multiple control planes and toolchains. Without a properly designed connectivity layer, scaling Kubernetes across clusters and clouds results in higher cost, slower troubleshooting, and reduced reliability.
The easiest way to run Kubernetes at the edge is to use a lightweight, consistent Kubernetes distribution combined with centralized networking and security control. Edge environments require predictable performance, components that are lightweight, minimal operational overhead, and support for intermittent connectivity. By standardizing ingress, traffic routing, and security policies across core, cloud, and edge, teams can deploy Kubernetes at scale while maintaining visibility, isolation, and automation without relying on cloud-specific services.
CNFs run as containerized services managed by Kubernetes rather than packaging network functions inside virtual machines. This enables Kubernetes-native automation, horizontal scaling, more granular resource allocation, and independent lifecycle management while reducing the infrastructure overhead associated with VM-based deployments.