TY - GEN

T1 - HyperAlloc

T2 - 20th European Conference on Computer Systems (EuroSys 2025) co-located with 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS 2025)

AU - Wrenger, Lars

AU - Albes, Kenny

AU - Wurps, Marco

AU - Dietrich, Christian

AU - Lohmann, Daniel

N1 - Publisher Copyright: © 2025 Copyright held by the owner/author(s).

PY - 2025/3/30

Y1 - 2025/3/30

N2 - The provisioning of the right amount of DRAM to virtual machines (VMs) is still a major challenge and cost driver in virtualization settings. Many VMs run applications with highly volatile memory demands, which either leads to massive overprovisioning in low-demand phases or poor QoS in high-demand phases. Memory hotplugging and ballooning have become established techniques (in Linux/KVM available via virtio-mem and virtio-balloon) to dynamically de/inflate the physical memory of a VM cooperatively, by having the guests give back unused memory to the hypervisor. However, current VM deflation techniques are either not DMA-safe, preventing the passthrough of important devices like GPUs or NICs, or are not flexible or fast enough to cope with the frequently changing demands of the guest. We present HyperAlloc, a DMA-safe and extremely efficient mechanism for virtual machine de/inflation. The core idea is to provide the hypervisor direct access to the guest’s page-frame allocator, greatly reducing the communication overhead. HyperAlloc can shrink virtual machines 362 times faster than virtio-balloon and 10 times faster than virtio-mem while having no measurable impact on the guest’s performance. HyperAlloc’s automatic reclamation provides for better memory elasticity by reducing the average memory footprint of a clang compilation by 17 percent compared to virtio-balloon’s free-page reporting while, again, having no measurable impact on the guest’s performance.

AB - The provisioning of the right amount of DRAM to virtual machines (VMs) is still a major challenge and cost driver in virtualization settings. Many VMs run applications with highly volatile memory demands, which either leads to massive overprovisioning in low-demand phases or poor QoS in high-demand phases. Memory hotplugging and ballooning have become established techniques (in Linux/KVM available via virtio-mem and virtio-balloon) to dynamically de/inflate the physical memory of a VM cooperatively, by having the guests give back unused memory to the hypervisor. However, current VM deflation techniques are either not DMA-safe, preventing the passthrough of important devices like GPUs or NICs, or are not flexible or fast enough to cope with the frequently changing demands of the guest. We present HyperAlloc, a DMA-safe and extremely efficient mechanism for virtual machine de/inflation. The core idea is to provide the hypervisor direct access to the guest’s page-frame allocator, greatly reducing the communication overhead. HyperAlloc can shrink virtual machines 362 times faster than virtio-balloon and 10 times faster than virtio-mem while having no measurable impact on the guest’s performance. HyperAlloc’s automatic reclamation provides for better memory elasticity by reducing the average memory footprint of a clang compilation by 17 percent compared to virtio-balloon’s free-page reporting while, again, having no measurable impact on the guest’s performance.

KW - Allocators

KW - Ballooning

KW - Overcommitment

KW - Virtual Machines

UR - http://www.scopus.com/inward/record.url?scp=105002248316&partnerID=8YFLogxK

U2 - 10.1145/3689031.3717484

DO - 10.1145/3689031.3717484

M3 - Conference contribution

AN - SCOPUS:105002248316

SP - 702

EP - 719

BT - EuroSys 2025 - Proceedings of the 2025 20th European Conference on Computer Systems

Y2 - 30 March 2025 through 3 April 2025

ER -