Efficient suspend and resume of instances

• US 10,353,731 B2
• Filed: 05/08/2017
• Issued: 07/16/2019
• Est. Priority Date: 06/08/2015
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Roth, Gregory Branchek
• Primary Examiner(s): Nguyen, Phillip H
• Application Number:
  US15/589,277
• Publication Number:
  US 20170242725A1
• Time in Patent Office: 799 Days
• Field of Search: 718 1, 718104, 718105
• CPC Class Codes: G06F 2009/45575 Starting, stopping, suspending or resuming virtual machine instances
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/461 Saving or restoring of program or task context
• CPC Class Codes: G06F 9/4868 with creation or replication
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: G06F 9/52 Program synchronisation; Mutual exclusion, e.g. by means of semaphores

Abstract

A plurality of virtual machines are each restored to a previous state, by, for each virtual machine of the plurality, obtaining a first snapshot of an exemplary virtual machine, obtaining a set of local changes of a virtual machine, deriving a snapshot of the virtual machine based at least in part on the first snapshot and the set of local changes, and restoring the virtual machine to a previous state based on the snapshot.

Claims

1. A computer-implemented method, comprising:

as a result of receiving a request to instantiate a second plurality of virtual machines to a previous state of a terminated first plurality of virtual machines;

obtaining a first snapshot of a first state common to the first plurality of virtual machines;

deriving a second snapshot of a second state of an individual virtual machine of the first plurality of virtual machines, the second snapshot derived based at least on the first snapshot and a first set of changes resulting from individual operation of the individual virtual machine;

deriving a third snapshot of the individual virtual machine based at least on the second snapshot and a second set of changes resulting from operation of the individual virtual machine from the second state to a third state; and

instantiating the second plurality of virtual machines to the previous state of the terminated first plurality of virtual machines based at least on the third snapshot, provisioning the second plurality of virtual machines.

2. The computer-implemented method of claim 1, wherein the first snapshot is a baseline snapshot of an exemplary virtual machine, the first snapshot used to instantiate the individual virtual machine.

3. The computer-implemented method of claim 1, wherein at least one of the first snapshot, second snapshot, and third snapshot includes contents of a physical memory of a computing system hosting the virtual machine.

4. The computer-implemented method of claim 1, wherein the request is a request by a scaling service to increase a total number of virtual machines in a fleet of virtual machines.

5. The computer-implemented method of claim 1, wherein the request is a request by a scaling service to resume operations of at least one of the virtual machines in a fleet of virtual machines following a previous suspension of operations of at least one of the virtual machines in the fleet of virtual machines.

6. A system, comprising memory to store instructions that, as a result of execution by one or more processors, cause the system to:

in response to a request to instantiate a virtual machine to a previous state of a terminated virtual machine, deriving a third snapshot of a different virtual machine of a set of virtual machines based at least on a first set of changes associated with a first snapshot of a first state of the different virtual machine and resulting from operation of the different virtual machine from the first state to a second state and a second set of changes associated with a second snapshot of the second state of the different virtual machine and resulting from operation of the different virtual machine from the second state to a third state, wherein the first snapshot is common to the set of virtual machines and the second set of changes is specific to the different virtual machine; and

instantiate the virtual machine to the previous state of the terminated virtual machine using the third snapshot and provisioning the virtual machine.

7. The system of claim 6, wherein the virtual machine being instantiated is a second virtual machine instance different from a virtual machine instance used to derive the third snapshot.

8. The system of claim 7, wherein the instructions further comprise instructions that, as a result of execution by the one or more processors, cause the system to terminate a virtual machine instance used to derive the third snapshot following the instantiation of the second virtual machine instance.

9. The system of claim 6, wherein the first set of changes comprises changes made since the first snapshot and the second set of changes comprises changes made since the second snapshot.

10. The system of claim 6, wherein the instantiation of the virtual machine is in response to a request to instantiate a virtual machine made by a virtual machine manager.

11. The system of claim 10, wherein the virtual machine manager is a hypervisor of a host computing system.

12. The system of claim 8, wherein the virtual machine used to derive the third snapshot is terminated because it indicates a lower priority than a third virtual machine competing for a set of resources, and wherein the second virtual machine represents a resumption of the state of the virtual machine used to derive the third snapshot after the third virtual machine releases the set of resources.

13. The system of claim 6, wherein the first snapshot is a baseline snapshot used to instantiate multiple virtual machines.

14. A non-transitory computer-readable storage medium comprising executable instructions that, if executed by one or more processors of a computer system, cause the computer system to at least:

in response to a request to instantiate a second virtual machine to a previous state of a terminated virtual machine, derive a second snapshot of a second state of a first virtual machine of a plurality of virtual machines at a second time, the second snapshot derived at least from a first snapshot of a first state of the first virtual machine at a first time and a first set of changes resulting from operation of the first virtual machine from the first state to the second state between the first time and the second time, wherein the first snapshot is of a first state of the first virtual machine and shared between the plurality of virtual machines, a respective first set of changes corresponding to each of the plurality of virtual machines;

derive, at a third time, a third snapshot at least from the second snapshot and a second set of changes resulting from operation of the first virtual machine from the second state to a third state between the second time and the third time;

terminate, after deriving the third snapshot, a virtual machine of the plurality of virtual machines; and

instantiate, after terminating and de-provisioning the virtual machine, the second virtual machine to the previous state of the terminated virtual machine based on the third snapshot and provisioning the second virtual machine.

15. The non-transitory computer-readable storage medium of claim 14, wherein instantiating the second virtual machine is triggered based at least on a variable cost value for a set of virtual machines passing a threshold price.

16. The non-transitory computer-readable storage medium of claim 14, wherein the second virtual machine being instantiated is a second virtual machine instance different from a virtual machine instance used to derive the third snapshot.

17. The non-transitory computer-readable storage medium of claim 16, wherein the virtual machine instance used to derive the third snapshot is terminated following the instantiation of the second virtual machine instance.

18. The non-transitory computer-readable storage medium of claim 16, wherein the virtual machine instance used to derive the third snapshot is not terminated following the instantiation of the second virtual machine instance.

19. The non-transitory computer-readable storage medium of claim 16, wherein a physical computing system executing the virtual machine instance used to derive the third snapshot is different than a physical computing system executing the second virtual machine instance.

20. The non-transitory computer-readable storage medium of claim 14, wherein the virtual machine was instantiated from a common software image containing contents and structure of a storage volume.