Designing VCF 9.1: Architecture, Sizing, and Scale in Practice

In this episode of the Virtually Speaking Podcast, hosts Pete Fletcher and John Nicholson welcome Ahmad Yunus, a product manager at Broadcom, to dive deep into the architectural, sizing, and scaling considerations for VMware Cloud Foundation 9.1. The conversation centers on practical design decisions that practitioners face when deploying or migrating to VCF, emphasizing the importance of understanding latency, concurrency, and component-level limits. Ahmad explains how VCF 9.1 introduces greater flexibility through features like the new VCF Ops platform, which allows customers to adopt capabilities incrementally—'crawl, walk, run'—without needing to implement full fleet or self-service functionality from day one. He highlights key improvements such as centralized documentation on the ConfigMax and Ports websites, including exportable JSON, CSV, and PDF formats for firewall and network planning, and the introduction of a unified sizing framework based on host and VM counts (5,000 hosts per instance, 50,000 per fleet). The episode also covers strategic design choices like fleet placement, regional segmentation, and licensing server flexibility, all aimed at reducing complexity and enabling scalable, resilient deployments. The discussion underscores a shift from rigid, monolithic requirements to a more modular, adaptable approach in VCF 9.1. Ahmad stresses the importance of assessing current environments against ConfigMax limits before upgrading, understanding the impact of new services like certificate and password management in Ops, and planning for both latency and concurrency during patching and lifecycle management. The episode concludes with actionable advice: start small, validate network and port configurations early, leverage centralized documentation, and build capabilities over time. These insights are designed to help enterprises avoid common pitfalls and achieve smoother, more successful VCF deployments.