Multi-hypervisor virtual machines

• US 10,437,627 B2
• Filed: 10/23/2017
• Issued: 10/08/2019
• Est. Priority Date: 11/25/2014
• Status: Active Grant

• Current Assignee: The Research Foundation for The State University of New York (State University of New York)
• Original Assignee: The Research Foundation for The State University of New York (State University of New York)
• Inventors: Hu, Yaohui, Gopalan, Kartik
• Primary Examiner(s): Sun, Charlie
• Application Number:
  US15/790,751
• Publication Number:
  US 20180060107A1
• Time in Patent Office: 715 Days
• CPC Class Codes: G06F 2009/4557 Distribution of virtual machine instances; Migration and load balancing
• CPC Class Codes: G06F 2009/45579 I/O management, e.g. providing access to device drivers or storage
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

Standard nested virtualization allows a hypervisor to run other hypervisors as guests, i.e. a level-0 (L0) hypervisor can run multiple level-1 (L1) hypervisors, each of which can run multiple level-2 (L2) virtual machines (VMs), with each L2 VM is restricted to run on only one L1 hypervisor. Span provides a Multi-hypervisor VM in which a single VM can simultaneously run on multiple hypervisors, which permits a VM to benefit from different services provided by multiple hypervisors that co-exist on a single physical machine. Span allows (a) the memory footprint of the VM to be shared across two hypervisors, and (b) the responsibility for CPU and I/O scheduling to be distributed among the two hypervisors. Span VMs can achieve performance comparable to traditional (single-hypervisor) nested VMs for common benchmarks.

Claims

1. A computer system, comprising:

at least one hardware processor;

a privileged hypervisor, executing on the at least one hardware processor;

a plurality of nested hypervisors, configured to concurrently execute under the privileged hypervisor each having a respective set of privileges lower than a set of privileges of the privileged hypervisor; and

at least one virtual machine, configured to concurrently communicate with the plurality of nested hypervisors, and having a memory image which is shared between at least two of the plurality of nested hypervisors, wherein each of the plurality of nested hypervisors provides at least one service or resource to the at least one virtual machine which is different from another of the plurality of nested hypervisors.

2. The computer system according to claim 1, wherein at least two of the plurality of nested hypervisors have respectively different sets of execution privileges.

3. The computer system according to claim 1, wherein the existence of the plurality of nested hypervisors is transparent to a guest operating system which executes on the at least one virtual machine.

4. The computer system according to claim 1, wherein coordination of access by the at least one virtual machine to the plurality of nested hypervisors is automatic, and an operating system and applications of the at least one virtual machine execute substantially without explicit control over a selection of respective ones of the plurality of nested hypervisors.

5. The computer system according to claim 1, wherein a single nested hypervisor has exclusive control over at least one portion of a memory map associated with the at least one virtual machine.

6. The computer system according to claim 1, wherein at least one respective virtual central processing unit of the at least one virtual machine executes under the plurality of nested hypervisors, and the plurality of nested hypervisors partition responsibility for scheduling the at least one respective virtual central processing unit.

7. The computer system according to claim 1, wherein input-output scheduling for a first of the plurality of nested hypervisors is handled by a second of the plurality of nested hypervisors.

8. The computer system according to claim 1, wherein the at least one virtual machine comprises a plurality of virtual machines, which each rely on at least two concurrently available nested hypervisors to interface with the physical host system.

9. The computer system according to claim 1, wherein the at least one virtual machine concurrently communicates directly with at least one of the plurality of nested hypervisors and the privileged hypervisor.

10. The computer system according to claim 1, wherein a responsibility for servicing input/output requests of the at least one virtual machine is partitioned at a device level granularity among the plurality of nested hypervisors.

11. The computer system according to claim 1, wherein the at least one virtual machine is configured to execute a guest operating system which supports a polling mode driver for receiving communications via at least one of the plurality of nested hypervisors, substantially without interrupts.

12. The computer system according to claim 1, wherein a first virtual central processing unit is controlled by a respective first nested hypervisor, and is configured to select a single nested hypervisor for relaying input/output requests to a second virtual central processing unit controlled by a second nested hypervisor.

13. The computer system according to claim 1, wherein a first virtual central processing unit is controlled by a respective first nested hypervisor, and is configured to select a single nested hypervisor for relaying device interrupts to a second virtual central processing unit controlled by a second nested hypervisor.

14. A method of operating a computer system, comprising:

providing a privileged hypervisor, executing on at least one hardware processor;

providing a plurality of nested hypervisors, executing under the privileged hypervisor, the plurality of nested hypervisors having a set of privileges lower than a set of privileges of the privileged hypervisor, at least two of the nested hypervisors being configured to-concurrently execute independently of each other;

providing at least one virtual machine, having a memory image shared between at least two of the plurality of nested hypervisors; and

concurrently communicating between the at least one virtual machine and the plurality of nested hypervisors, wherein each of the plurality of nested hypervisors provides at least one service or resource different from another of the plurality of nested hypervisors to the at least one virtual machine.

15. The method according to claim 14, wherein at least two of the plurality of nested hypervisors have respectively different sets of execution privileges.

16. The method according to claim 14, further comprising automatically coordinating concurrent access by the at least one virtual machine to the plurality of nested hypervisors, wherein an operating system and applications of the at least one virtual machine execute substantially without explicit control over a selection of the plurality of nested hypervisors.

17. The method according to claim 14, further comprising exclusively controlling at least one portion of a memory map associated with the at least one virtual machine with a single nested hypervisor.

18. The method according to claim 14, further comprising:

executing at least one respective virtual central processing unit of the virtual machine under the plurality of nested hypervisors; and

partitioning responsibility for scheduling the at least one respective virtual central processing unit by the plurality of nested hypervisors.

19. The method according to claim 14, further comprising controlling at least one of an input-output operation and an interrupt operation by a first of the plurality of nested hypervisors on behalf of a second of the plurality of nested hypervisors.

20. A method for providing multiple hypervisors for a virtual machine, comprising:

providing a virtual machine executing on hardware resources of a unitary host machine under at least two concurrently available and independently executing nested hypervisors which interface the virtual machine to the unitary host machine through a privileged hypervisor, wherein the at least two concurrently available and independently executing nested hypervisors share a memory image of the virtual machine, and access to at least a respective hardware resource; and

performing a series of actions by the virtual machine which employ the respective hardware resource selectively provided by respective ones of the at least two concurrently available and independently executing nested hypervisors without mutual interference.