Scaling and Traffic-Managed Model Context Protocol (MCP) with BIG-IP Next for K8s
Introduction As AI models get more advanced, running them at scale—especially in cloud-native environments like Kubernetes—can be tricky. That’s where the Model Context Protocol (MCP) comes in. MCP makes it easier to connect and interact with AI models, but managing all the traffic and scaling these services as demand grows is a whole different challenge. In this article and demo video, I will show how F5's BIG-IP Next for K8s (BNK), a powerful cloud native traffic management platform from F5 can solve that and keep things running smoothly and scale your MCP services as needed. Model Context Protocol (MCP) in a nutshell. There were many articles explaining what is MCP on the internet. Please refer to those in details. In a nutshell, it is a standard framework or specification to securely connect AI apps to your critical data, tools, and workflow. The specification allow Tracking of context across multiple conversation Tool integration — model call external tools Share memory/state — remember information. MCP’s "glue" model to tools through a universal interface "USB-C for AI" What EXACTLY does MCP solve? MCP addresses many challenges in the AI ecosystem. I believe two key challenges it solve Complexities of integrating AI Model (LLM) with external sources and tools By standardization with a universal connector ("USB-C for AI") Everyone build "USB-C for AI" port so that it can be easily plug in each other Interoperability. Security with external integration Framework to establish secure connection Managing permission and authorization. What is BIG-IP’s Next for K8s (BNK)? BNK is F5 modernized version of the well-known Big-IP platform, redesigned to work seamlessly in cloud-native environments like Kubernetes. It is a scalable networking and security solution for ingress and egress traffic control. It builds on decades of F5's leadership in application delivery and security. It powers Kubernetes networking for today's complex workloads. BNK can be deployed on X86 architecture or ARM architecture - Nvidia Data Processing Unit (DPU) Lets see how F5's BNK scale and traffic managed an AIOps ecosystem. DEMO Architecture Setup Video Key Takeaways BIGIP Next for K8s, the backbone of the MCP architecture Technology built on decades of market-leading application delivery controller technology Secure, Deliver, and Optimize your AI infrastructure Provides deep insight through observability and visibility of your MCP traffic.
Agentic RAG - Securing GenAI with F5 Distributed Cloud Services
Agentic RAG (Retrieval-Augmented Generation) enhances the capabilities of a GenAI chatbot by integrating dynamic knowledge retrieval into its conversational abilities, making it more context-aware and accurate. In this demo, I will focus on security aspect of the solution. This demonstration will highlight the various security measures implemented and enforced in our AI reference architecture for this Agentic RAG. F5 is a trusted leader in security, with a track record of delivering robust solutions for securing applications and networks. Recognized by many independent evaluations as a Leader in Web Application and API Security from IDC, SC Award, TrustRadius, EMA, and many more, F5 exemplifies excellence and innovation. These endorsements affirm F5’s expertise, reassuring organizations that their digital assets are protected by a capable, reputable partner that keeps pace with evolving security needs.
Adding variable to generic Message Box
I'm having issues with my AD Query. The AD Query agent expression is as such: expr { [mcget {session.ad.last.attr.memberOf}] contains "Test.40.Employee" && [mcget {session.ad.last.authresult}] == 1 } The AD Query isn't sending the user agents down the expected branch, so I'm trying to carry the defined variable into a generic message box to troubleshoot. In the generic message box, I added the below value: AD Query variable %{session.ad.last.attr.memberOf} However when I go through the VPE, the value is not displayed in the message box. Any help on where I'm going wrong?
Installing McAfee Agent for SSL Decryption at F5
McAfee agent can be installed in the Linux Based OS. Even though F5 is a CentOS-based Linux installation, I understand that the McAfee agent cannot be installed on it. I am looking for any supporting documentation published by F5 which says that external agents cant be installed on it.
Can't get F5 Lbaas agent run on Openstack.
hi, I'm new in Openstack and F5. My openstack is Havana, and I have contoller node, compute node, network node running in environment. When I install f5-lbaas-driver in both controller and network node, it seems good. But, when I install f5-bigip-lbaas-agent in network node, problem comes. I can't find "Loadbalance agent" entry with command "neutron agent-list". in log file "/var/log/neutron/f5-bigip-lbaas-agent.log", "ERROR neutron.common.legacy [-] Skipping unknown group key: firewall_driver" shows a lot. Meanwhile, I see this ERROR message in console, "Error importing loadbalancer device driver: neutron.services.loadbalancer.drivers.f5.bigip.icontrol_driver.iControlDriver". It makes me crazy. The agent file is "f5-bigip-lbaas-agent_1.0.1.havana-1-1_all.deb". Can anybody help me solve this problem? Thanks so much.
