FreeBSD bhyve intro
summary
A minimal FreeBSD hypervisor originally written 2011 and upstreamed later, bhyve runs as a UNIX process with vCPUs as threads, uses a kernel module + character device + lib + bhyve process and bhyvecontrol, and requires CPUs with EPT (second-level page tables) (Nehalem+). Initial tree supported up to eight vCPUs, paravirtual I/O, and boot via UEFI; today it offers PCI pass-through, a built‑in VNC framebuffer, demand‑paged guest memory, broad BSD/Linux guest support and limited recent Windows 64‑bit support. Lacks nested virtualization, GUI front end, GPU passthrough and live migration.
AI-written summary · about
chapters (11)
- 0:00Intro
- 2:02bhyve history
- 9:40Supported guests
- 11:12Components
- 12:26Controversy
- 13:47Running Windows
- 16:38Future work
- 19:57Questions
- 22:47Challenges
- 28:06Making it easier
- 29:42GPU passthrough
related talks
24mLightweight VMs for serverless and containers
18mUnikernels for Serverless Computing
20mFirecracker: Lightweight Virtualization for Serverless Applications
20mExpeditious High-Concurrency MicroVM SnapStart in Persistent Memory with an...
44mUnikernels and Hyper Elastic Clouds
46mIncludeOS - running C++ applications without an operating system
05mWhat's Exciting Now In Bootc, And What's Next?
44mThe Containerization of the Operating System: Exploring Bootc and the Future of Linux Deployment
