Dependency handling in an on-demand network code execution system

• US 10,564,946 B1
• Filed: 12/13/2017
• Issued: 02/18/2020
• Est. Priority Date: 12/13/2017
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Wagner, Timothy Allen, Moffatt, Bryan Nicholas, Brooker, Marc John, Rodrigues dos Santos, Robison, Jayakar, Niranjan
• Primary Examiner(s): Nguyen, Duy Khuong T
• Application Number:
  US15/841,143
• Time in Patent Office: 797 Days
• CPC Class Codes: G06F 2009/45562 Creating, deleting, cloning virtual machine instances
• CPC Class Codes: G06F 2009/45575 Starting, stopping, suspending or resuming virtual machine instances
• CPC Class Codes: G06F 8/433 Dependency analysis; Data or control flow analysis
• CPC Class Codes: G06F 8/443 Optimisation
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

Systems and methods are described for managing dependencies among code in an on-demand code execution system in a manner that enables efficient execution of the code. A user may generate a task on the system by submitting code, which may reference additional code on the system. The system can determine a dependency graph for the submitted code, and generate packages for the task based at least partly on the dependency graph. On receiving a request to execute the task, the system can provision an execution environment with the packages, and execute the code of the task. The packages may be selected based on analysis of the dependency graph, such that execution of the task on the on-demand code execution system is expected to satisfy a desired execution metric, which may include a maximum time to begin executing code of a task subsequent to receiving a request to execute the task.

Claims

1. A system comprising:

one or more non-transitory data stores configured to store code objects, individual code objects associated with an object identifier and including code executable on an on-demand code execution system; and

one or more processors in communication with the one or more non-transitory data stores and configured with computer-executable instructions to;

receive a request to generate a task on the on-demand code execution system, the request including a code object including task code executable to implement the task and an identifier of at least one additional code object required to execute the task code;

generate a dependency graph for the task, the dependency graph identifying the code object included within the request and the at least one additional code object corresponding to the identifier of the request;

group code objects identified in the dependency graph into one or more packages, individual packages of the one or more packages including at least one code object identified within the dependency graph, wherein the code objects are grouped into the one or more packages based at least partly on an expected time to provision an execution environment on the on-demand code execution system with the one or more packages, and wherein the expected time to provision the execution environment is based at least partly on an expected caching of individual packages within the one or more packages on the on-demand code execution system and an expected time to transfer uncached packages of the one or more packages between the one or more non-transitory data stores and the execution environment;

generate the one or more packages;

store the one or more package within the one or more non-transitory data stores;

receive a request to execute the task on the on-demand code execution system; and

execute the task on the on-demand code execution system at least partly by retrieving the one or more packages from the one or more non-transitory data stores and provisioning the execution environment of the on-demand code execution system with the one or more packages.

2. The system of claim 1, wherein the at least one additional code object corresponds to at least one of a code library, frameworks, software development kits (SDKs), drivers, or a runtime.

3. The system of claim 1, wherein the identifier of at least one additional code object required to execute the task code corresponds to a uniform resource identifier (URI).

4. The system of claim 1, wherein generating the dependency graph for the task comprises identifying, within the at least one additional code object corresponding to the identifier of the request, a reference to a third code object, and wherein the dependency graph identifies the third code object as a dependency of the at least one additional code object.

5. A computer-implemented method comprising:

receiving a request to generate a task on an on-demand code execution system, the request including a code object including task code executable to implement the task and an identifier of at least one additional code object required to execute the task code;

generating a dependency graph for the task, the dependency graph identifying the code object included within the request and the at least one additional code object corresponding to the identifier of the request;

grouping code objects identified in the dependency graph into one or more packages, individual packages of the one or more packages including at least one code object identified within the dependency graph, wherein the code objects are grouped into the one or more packages based at least partly on an expected time to provision an execution environment on the on-demand code execution system with the one or more packages, and wherein the expected time to provision the execution environment is based at least partly on an expected caching of individual packages within the one or more packages on the on-demand code execution system and an expected time to transfer uncached packages of the one or more packages to the execution environment;

generating the one or more packages;

receiving a request to execute the task on the on-demand code execution system; and

executing the task on the on-demand code execution system at least partly by provisioning the execution environment of the on-demand code execution system with the one or more packages.

6. The computer-implemented method of claim 5, wherein the at least one additional code object corresponding to the identifier of the request is associated with a set of permission requirements for utilizing the at least one additional code object, wherein the request is received from a user associated with a set of user permissions, and wherein the computer-implemented method further comprises verifying that the set of user permissions satisfies the a set of permission requirements.

7. The computer-implemented method of claim 5, wherein individual packages of the one or more packages correspond to individual files, and wherein at least one of the individual packages corresponds to a compressed file system image file.

8. The computer-implemented method of claim 5, wherein the expected caching of individual packages within the one or more packages on the on-demand code execution system is based on an expected frequency of use of the individual packages on the on-demand code execution system.

9. The computer-implemented method of claim 5 further comprising determining that the expected time to provision an execution environment on the on-demand code execution system with the one or more packages satisfies a threshold value.

10. Non-transitory computer readable media including computer-executable instructions that, when executed by a processor of an on-demand code execution system, cause the processor to:

receive a request to generate a task on the on-demand code execution system, the request including a code object including task code executable to implement the task and an identifier of at least one additional code object required to execute the task code;

generate a dependency graph for the task, the dependency graph identifying the code object included within the request and the at least one additional code object corresponding to the identifier of the request;

group code objects identified in the dependency graph into one or more packages, individual packages of the one or more packages including at least one code object identified within the dependency graph, wherein the code objects are grouped into the one or more packages based at least partly on an expected time to provision an execution environment on the on-demand code execution system with the one or more packages, and wherein the expected time to provision the execution environment is based at least partly on an expected caching of individual packages within the one or more packages on the on-demand code execution system and an expected time to transfer uncached packages of the one or more packages to the execution environment;

generate the one or more packages;

receive a request to execute the task on the on-demand code execution system; and

execute the task on the on-demand code execution system at least partly by provisioning the execution environment of the on-demand code execution system with the one or more packages.

11. The non-transitory computer readable media of claim 10, wherein the dependency graph includes a plurality of nodes representing individual code objects, and wherein the dependency graph further includes edges between the plurality of nodes, the edges representing dependencies between the individual code objects.

12. The non-transitory computer readable media of claim 10, wherein generating the one or more packages comprises combining the code object included within the request with all code objects identified within the dependency graph as a dependency of the code object included within the request.

13. The non-transitory computer readable media of claim 10, wherein generating the one or more packages comprises combining the code object included within the request with only code objects identified within the dependency graph as a dependency of the code object included within the request and not a dependency of any other code object.

14. The non-transitory computer readable media of claim 10, wherein generating the one or more packages comprises combining the code object included within the request with all non-runtime code objects identified within the dependency graph as a dependency of the code object included within the request.

15. The non-transitory computer readable media of claim 10, wherein the one or more packages include plurality of packages, and wherein the computer-executable instructions further cause the processor to:

determine that an execution metric of the task does not satisfy a threshold value; and

determining a modification of the plurality of packages for the task expected to enable the execution metric to subsequently satisfy the threshold value; and

modify the plurality of packages according to the determined modification.

16. The non-transitory computer readable media of claim 15, wherein the execution metric is based on historical executions of the task.

17. The non-transitory computer readable media of claim 10, wherein the at least one additional code object corresponds to at least one of source code or machine code.