Execution locations for request-driven code

• US 9,830,449 B1
• Filed: 12/16/2015
• Issued: 11/28/2017
• Est. Priority Date: 12/16/2015
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Wagner, Timothy Allen
• Primary Examiner(s): Shayanfar, Ali
• Application Number:
  US14/971,859
• Time in Patent Office: 713 Days
• CPC Class Codes: G06F 2009/45587 Isolation or security of virtual machine instances
• CPC Class Codes: G06F 21/53 by executing in a restricted environment, e.g. sandbox or secure virtual machine
• CPC Class Codes: G06F 21/6281 at program execution time, where the protection is within the operating system
• CPC Class Codes: G06F 2221/034 Test or assess a computer or a system
• CPC Class Codes: G06F 9/45508 Runtime interpretation or emulation, e g. emulator loops, bytecode interpretation
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/5005 to service a request
• CPC Class Codes: G06F 9/5077 Logical partitioning of resources; Management or configuration of virtualized resources specific details on emulation or internal functioning of virtual machines G06F9/455

Abstract

Systems and methods are described for determining a location in an on-demand code execution environment to execute user-specified code. Virtual machines can be utilized to isolate different executions of code from one another. However, virtual machines require additional computing resources to implement, and may slow code executions. This disclosure enables multiple items of code, potentially associated with different users, to be executed on a single virtual machine instance or other device. Specifically, the present disclosure enables the generation of a risk profile for an item of code, which can be used to determine the security or privacy risk that would occur by executing the code on a device. By comparing the risk profiles of a given item of code to other items of code executing on a device, the on-demand code execution environment can selectively group code executions, thus increasing the efficiency of the system while maintaining security and privacy.

Claims

1. A computer-implemented method of selecting virtual machine instances on which to execute user-specified code within an on-demand code execution environment, the on-demand code execution environment comprising a plurality of virtual machine instances executing an initialized operating system and awaiting requests to execute user-specified code within that operating system, the method comprising:

as implemented by one or more computing devices configured with specific computer-executable instructions,

generating a first risk profile for a first task associated with the on-demand code execution environment, the first task corresponding to a first set of user-defined executable code, wherein the risk profile for the first task indicates a risk that execution of the first task on a virtual machine instance will compromise security of other tasks executing on the virtual machine instance, wherein the first risk profile for the first task is determined based at least in part on at least one of one or more function calls included within the first set of user-defined code, one or more libraries included within the first set of user-defined code, one or more computing resources expected to be utilized during execution of the first set of user-defined code, or one or more permissions requested for execution of the first set of user-defined code;

obtaining a request to execute the first task on the on-demand code execution environment;

selecting the virtual machine instance from the plurality of virtual machine instances on which to execute the first task at least partly by;

obtaining a second risk profile for a second task assigned to be executed on the virtual machine instance, the second task corresponding to a second set of user-defined executable code, wherein the risk profile for the second task indicates a risk that execution of the second task on the virtual machine instance will compromise security of other tasks executing on the virtual machine instance; and

determining, based at least in part on the first risk profile and the second risk profile, that a risk level associated with execution of both the first task and second task on the virtual machine instance does not exceed a threshold value; and

assigning the virtual machine instance to execute the first task.

2. The computer-implemented method of claim 1, wherein the risk profile for the first task indicates which of a set of computing resources are expected to be utilized during execution of the first task, and wherein the risk profile for the second task indicates which of the set of computing resources are expected to be utilized during execution of the second task.

3. The computer-implemented method of claim 2, wherein determining that the risk level associated with execution of both the first task and second task on the virtual machine instance does not exceed the threshold value comprises determining that at least one computing resources of the set of computing resources is unlikely to be utilized during executions of both the first and second tasks.

4. The computer-implemented method of claim 1, wherein the risk profile for the first task is determined based at least in part on at least one of static analysis of the first set of user-defined executable code or dynamic analysis of executions of the a first set of user-defined executable code.

5. The computer-implemented method of claim 1, wherein the computing resources expected to be utilized during execution of the first set of user-defined code include at least one of computing resources local to the virtual machine instance or computing resources external to the virtual machine instance.

6. The computer-implemented method of claim 1, further comprising:

obtaining a request to execute a third task on the on-demand code execution environment;

obtaining a third risk profile for the third task, wherein the third risk profile for the third task indicates a risk that execution of the third task on the virtual machine instance will compromise security of other tasks executing on the virtual machine instance; and

determining, based at least in part on the third risk profile, that a risk level associated with execution of the third task on the virtual machine instance exceeds a threshold value; and

declining to execute the third task on the virtual machine instance.

7. A system for selecting devices on which to execute user-specified code within an on-demand code execution environment, the system comprising:

a non-transitory data store including a risk profile for a task associated with the on-demand code execution environment, the task corresponding to a first set of executable code that may be executed by a device of the on-demand code execution environment, wherein the risk profile for the task indicates a risk that execution of the task on the device will compromise security of other processes executing on the device, and wherein the risk profile for the task is determined based at least in part on at least one of one or more function calls included within the first set of executable code, one or more libraries included within the first set of executable code, one or more computing resources expected to be utilized during execution of the first set of executable code, or one or more permissions requested for execution of the first set of executable code; and

a computing device comprising a processor in communication with the non-transitory data store and configured with specific computer-executable instructions to;

obtain a request to execute the task on the on-demand code execution environment;

obtain a second risk profile for a process assigned to be executed on the device, wherein the risk profile for the process indicates a risk that execution of the process on the device will compromise security of other processes executing on the device;

determine, based at least in part on the first risk profile and the second risk profile, that a risk level associated with execution of both the task and the process on the device does not exceed a threshold value; and

assign the device to execute the first task.

8. The system of claim 7, wherein the process is a second task defined by a user of the on-demand code execution environment.

9. The system of claim 7, wherein the process is associated with an auxiliary service.

10. The system of claim 7, wherein the device is at least one of a virtual machine instance or a physical computing device.

11. The system of claim 7, wherein the risk profile for the task indicates a risk level of the task in a plurality of risk categories, and wherein determining that the risk level associated with execution of both the task and the process on the device does not exceed a threshold value comprising determining that a combined risk level of the task and the process in any of the plurality of risk categories does not exceed a threshold value.

12. The system of claim 7, wherein the specific computer-executable instructions further configure the computing device to:

obtain a request to execute a second task on the on-demand code execution environment;

obtain a risk profile for the second task, wherein the risk profile for the second task indicates a risk that execution of the second task on the device will compromise security of other processes executing on the device; and

determine, based at least in part on the third risk profile, that a risk level associated with execution of the third task on the device exceeds a threshold value; and

decline processing of the second task on the on-demand code execution environment.

13. The system of claim 7, wherein the specific computer-executable instructions further configure the computing device, subsequent to declining processing of the second task on the on-demand code execution environment, to regenerate the request to execute the second task on the on-demand code execution environment.

14. The system of claim 13, wherein the specific computer-executable instructions further configure the computing device to impose a delay prior to regenerating the request to execute the second task on the on-demand code execution environment.

15. The system of claim 13, wherein the specific computer-executable instructions further configure the computing device to transmit the regenerated request to a distinct computing system associated with the on-demand code execution environment.

16. Non-transitory, computer-readable storage media comprising computer-executable instructions for selecting devices on which to execute user-specified code within an on-demand code execution environment, wherein the computer-executable instructions, when executed by a computer system, cause the computer system to:

generate a risk profile for a process executing on a device of the on-demand code execution environment, wherein the risk profile for the process indicates a risk that execution of the process on the device will compromise security of other processes executing on the device, and wherein the risk profile for the device is determined based at least in part on at least one of one or more function calls included within code corresponding to the process, one or more libraries included within the code, one or more computing resources expected to be utilized during execution of the process, or one or more permissions requested for execution of the process;

obtain a request to execute a task on the on-demand code execution environment;

obtain a risk profile for the task, wherein the risk profile for the task indicates a risk that execution of the task on the device will compromise security of other processes executing on the device;

determine, based at least in part on the risk profile for the process and the risk profile for the task, that a risk level associated with execution of both the task and the process on the device does not exceed a threshold value; and

assign the device to execute the task.

17. The non-transitory, computer-readable storage media of claim 16, wherein the risk threshold for the device is determined by combining risk levels associated with individual processes of the other processes.

18. The non-transitory, computer-readable storage media of claim 17, wherein the risk levels associated with individual processes of the other processes are determined based on risk profiles associated with the individual processes.

19. The non-transitory, computer-readable storage media of claim 16, wherein the risk level associated with execution of both the task and the process on the device is determined not to exceed the threshold value based at least in part on determining that execution of the task is unlikely to utilize at least one computing resource of the device that is also utilized by the process.

20. The non-transitory, computer-readable storage media of claim 16, wherein the risk profile for the task is determined based at least in part on at least one of static analysis of the a set of user-defined executable code corresponding to the task or dynamic analysis of executions of the set of user-defined executable code.

21. The non-transitory, computer-readable storage media of claim 16, wherein the risk profile for the task is determined based at least in part on at least one of function calls included within a set of user-defined executable code corresponding to the task, libraries included within the set of user-defined code, computing resources expected to be utilized during execution of the set of user-defined code, or permissions requested for execution of the set of user-defined code.