Learn Cloud Hypervisor (a modern Rust-based VMM for cloud workloads) from the basics to production-grade: pre-requisites skill & environment setup, history & background & why you need it, core concepts & main architecture, installation & build, first Linux & UEFI boot, CPU memory & machine config, storage block pmem & virtio-fs, networking tap vhost-user & vDPA, hotplug CPU memory & device, cloud images & cloud-init, snapshot & restore, live migration, virtual I/O virtio IOMMU & security, network security & isolation, confidential computing TDX & SGX, Windows guest & UEFI, testing & CI, v53.0 & roadmap, Kata Containers kata-clh backend, nested virtualization & EC2, performance & observability, up to the ecosystem & final reflection, with a total of 23 episodes.
Before touching Cloud Hypervisor, you need to master basic Linux administration (KVM, CLI, TAP/bridge networking), understand the concepts of VMM and guest OS, and prepare an x86_64/aarch64 host with KVM, a kernel, a cloud image, and the correct firmware. In this episode you'll also install the Cloud Hypervisor binary and verify that the entire environment is ready.

This episode reviews the history of Cloud Hypervisor from Intel's 2019 initiative to open governance with AMD, Arm, and Microsoft, including the first LTS release v28. You'll also understand why this Rust VMM is needed: fast boot, small footprint, hotplug, snapshot/restore, live migration, and its role as a Kata Containers backend.

This episode dissects the Cloud Hypervisor architecture: a Rust single-binary VMM on top of the KVM/MSHV backend, virtio devices (net, block, pmem, fs, vsock, rng), and vhost-user offload. You'll also understand the role of supporting components such as rust-hypervisor-firmware, edk2 UEFI (CLOUDHV.fd), and the VM process flow from guest kernel to device model.

This episode guides you through installing Cloud Hypervisor two ways: downloading the v53.0 release binary or compiling from source with cargo build --release. You'll also set up setcap cap_net_admin+ep, verify the version, and make sure /dev/kvm and CPU virtualization support are ready to use.

This is the first moment your VM comes alive: direct boot with --kernel vmlinux and PVH, then firmware boot with rust-hypervisor-firmware and CLOUDHV.fd (edk2) for full OSes like Ubuntu. This episode dissects both paths, prepares the kernel and cloud image, and verifies that the VM console successfully displays kernel logs.

This episode covers VM resource configuration: --cpus boot=N with topology, features, and PMU; --memory with size, hotplug, and hugepages; and a comparison of CLI flags vs. a JSON config file via --config. You'll also learn shared memory and balloon tuning for efficiency.

This episode covers storage in Cloud Hypervisor: virtio-blk disks from various formats (raw, qcow2), persistent memory (pmem), and filesystem sharing with the host via virtio-fs. You'll also learn image conversion with qemu-img, building a multi-disk setup, and when to choose each storage type.

This episode builds VM networking: setting up TAP and bridge on the host, connecting a NIC to the guest with --net tap=...,mac=...,ip=..., offloading the data plane to a vhost-user daemon for high performance, and getting to know the still-experimental vDPA. You'll also learn how to build a multi-NIC setup.

This episode covers adding resources live: adding vCPUs and memory through the API socket without restarting, their reflection in the guest via ACPI, and VFIO device passthrough for GPUs and other devices. You'll also get to know the still-experimental vfio-user and safe practices for hotplugging in production.

This episode prepares VMs that are ready to use right away: downloading cloud images (Ubuntu, Fedora, Debian), converting them to raw, booting via direct or UEFI, then automating bootstrap with cloud-init (password, SSH key, and packages) through user-data and metadata.
