Orphaned VMs: Running VMs Uninterrupted While Host Kernel Is Offline For Reboots/Updates

Here's some very intriguing work taking place by a Google engineer... With the hyperscalers and other cloud providers pushing for zero-downtime, a set of experimental patches for "Orphaned VMs" allow for the virtual machines to continue to run while the host kernel is offline for security updates or rebooting with the Live Update Orchestrator. The experimental patches and request for feedback from other Linux kernel stakeholders was sent out this weekend on Orphaned VMs. Orphaned Virtual Machines are designed to run uninterrupted with continued execution on preserved physical CPUs across a host Linux kernel live update using the Live Update Orchestrator (LUO) that was spearheaded by Google. These RFC patches have been tested on Intel, AMD, and Arm server processors but still considered a "very early" work in progress and not production ready. This work was led by Google's Pasha Tatashin. This work is quite involved to get Orphaned VMs actually working when the host kernel goes down temporarily for an update with in-RAM vCPU state preservation, physical CPU preservation core support, an on-core scheduling framework, and the KVM Caretaker Core infrastructure. Those wishing to learn more can see this RFC patch series presently at 46 patches. "As cloud infrastructure continues to push toward zero-downtime host maintenance, extending the capabilities of kexec-based Live Update to minimize guest disruption is becoming increasingly critical. This proposal introduces the architectural concept of an "Orphaned VM"—a virtual machine that actively executes guest instructions on isolated physical hardware while completely decoupled from a host operating system or a VMM during a host reboot. This execution asymmetry is achieved by introducing the "Caretaker," a specialized bare-metal primitive interpose layer that runs within the hypervisor's privileged hardware state to trap and resolve some VM exits locally while the host kernel is offline. This session will explore the mechanisms required to orchestrate this transition using the LUO, focusing on how vcpufd structures are preserved across the reboot gap, how physical CPUs are shielded from boot-time reset signals, and how asynchronous hardware challenges like timekeeping drift, stray interrupts, and guest-to-guest IPI routing are mitigated during the management gap." The work is being discussed further next month at the Linux Plumbers Conference in Prague.