Distributed computing system node management

• US 10,243,870 B1
• Filed: 12/23/2014
• Issued: 03/26/2019
• Est. Priority Date: 12/23/2014
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Alton, Darren James, Zeng, Xiao, Villalobos, Adam Julio, Merritt, Eric Benjamin, Zgraggen, Guido Beat, Mhley, Christopher Ryan
• Primary Examiner(s): Survillo, Oleg
• Assistant Examiner(s): Moorad, Imran
• Application Number:
  US14/582,051
• Time in Patent Office: 1,554 Days
• Field of Search: 709224
• CPC Class Codes: A63F 13/352 involving special game server arrangements, e.g. regional servers connected to a national server or a plurality of servers managing partitions of the game world
• CPC Class Codes: G06F 21/44 Program or device authentication
• CPC Class Codes: G06F 21/554 involving event detection and direct action
• CPC Class Codes: G06F 2209/502 Proximity
• CPC Class Codes: G06F 2209/509 Offload
• CPC Class Codes: G06F 9/00 Arrangements for program control, e.g. control units program control for peripheral devices G06F13/10
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources

Abstract

Techniques for distributed computing system node management are described herein. In some cases, internal compute nodes (i.e., compute nodes that are allocated to the distributed system) may be mutually trusted such that they may freely establish communications with one another. By contrast, external compute nodes (i.e., compute nodes that aren'"'"'t allocated to the distributed computing system) may be untrusted such that their access to the distributed system may be regulated. In some cases, one or more of the compute nodes within the distributed computing system may maintain respective collections of system view information. Each respective collection of system view information may include, for example, information associated with the corresponding compute node'"'"'s view of the distributed computing system based on information that is available to the corresponding compute node.

Claims

1. A computer-implemented method for compute node allocation in a distributed computing system comprising a plurality of compute nodes comprising:

detecting an occurrence of a condition associated with allocation of at least one additional compute node to the distributed computing system;

determining an estimated probability associated with the allocation of the at least one additional compute node, wherein the estimated probability is determined based, at least in part, on a first likelihood that one or more other of the plurality of compute nodes will issue an allocation request, wherein the plurality of compute nodes execute on one or more computing devices;

performing a routine that returns either a first result or a second result, wherein returning of the first result or the second result is based on the estimated probability; and

issuing a request to allocate the at least one additional compute node to the distributed computing system based on the first result.

2. The computer-implemented method of claim 1, wherein the condition comprises determining that one or more of the plurality of compute nodes are exceeding a particular proportion of operation in a non-idle state.

3. The computer-implemented method of claim 1, wherein the estimated probability is determined based, in part, on a number of compute nodes desired for the allocation.

4. The computer-implemented method of claim 1, wherein the plurality of compute nodes collectively issue only a desired number of allocation requests.

5. The computer-implemented method of claim 1, wherein the estimated probability is determined based, in part, on a number of compute nodes within the plurality of compute nodes.

6. A computer-implemented method for compute node de-allocation in a distributed computing system comprising a plurality of compute nodes comprising:

detecting an occurrence of a condition associated with de-allocation of at least one compute node from the distributed computing system;

determining an estimated probability associated with the de-allocation of the at least one compute node, wherein the estimated probability is determined based, at least in part, on a first likelihood that one or more other of the plurality of compute nodes will issue a de-allocation request, wherein the plurality of compute nodes execute on one or more computing devices;

performing a routine that returns either a first result or a second result, wherein returning of the first result or the second result is based on the estimated probability; and

issuing a request to de-allocate the at least one compute node from the distributed computing system based on the first result.

7. The computer-implemented method of claim 6, wherein the condition comprises determining that one or more of the plurality of compute nodes are exceeding a particular proportion of operation in an idle state.

8. The computer-implemented method of claim 6, wherein the estimated probability is determined based, in part, on a number of compute nodes desired for the de-allocation.

9. The computer-implemented method of claim 6, wherein the plurality of compute nodes collectively issue only a desired number of de-allocation requests.

10. The computer-implemented method of claim 6, wherein the estimated probability is determined based, in part, on a number of compute nodes within the plurality of compute nodes.

11. A first compute node comprising:

one or more processors;

one or more memories having stored therein instructions that, upon execution by the one or more processors, cause the one or more processors to perform operations comprising;

detecting an occurrence of a condition associated with allocation of at least one additional compute node to a distributed computing system that comprises a plurality of compute nodes including the first compute node;

determining an estimated probability associated with the allocation of the at least one additional compute node, wherein the estimated probability is determined based, at least in part, on a first likelihood that one or more other of the plurality of compute nodes will issue an allocation request, wherein the plurality of compute nodes execute on one or more computing devices;

performing a routine that returns either a first result or a second result, wherein returning of the first result or the second result is based on the estimated probability; and

issuing a request to allocate the at least one additional compute node to the distributed computing system based on the first result.

12. The first compute node of claim 11, wherein the estimated probability is determined based, in part, on a number of compute nodes desired for the allocation.

13. The first compute node of claim 11, wherein the plurality of compute nodes collectively issue only a desired number of allocation requests.

14. The first compute node of claim 11, wherein the condition comprises determining that one or more of the plurality of compute nodes are exceeding a particular proportion of operation in a non-idle state.

15. The first compute node of claim 11, wherein the estimated probability is determined based, in part, on a number of compute nodes within the plurality of compute nodes.

16. One or more non-transitory computer-readable storage media having stored thereon instructions that, upon execution by a first compute node of a plurality of compute nodes, cause the first compute node to perform operations comprising:

detecting an occurrence of a condition associated with de-allocation of at least one compute node from a distributed computing system comprising the plurality of compute nodes;

determining an estimated probability associated with the de-allocation of the at least one compute node, wherein the estimated probability is determined based, at least in part, on a first likelihood that one or more other of the plurality of compute nodes will issue a de-allocation request, wherein the plurality of compute nodes execute on one or more computing devices;

performing a routine that returns either a first result or a second result, wherein returning of the first result or the second result is based on the estimated probability; and

issuing a request to de-allocate the at least one compute node from the distributed computing system based on the first result.

17. The one or more non-transitory computer-readable storage media of claim 16, wherein the condition comprises determining that one or more of the plurality of compute nodes are exceeding a particular proportion of operation in an idle state.

18. The one or more non-transitory computer-readable storage media of claim 16, wherein the estimated probability is determined based, in part, on a number of compute nodes desired for the de-allocation.

19. The one or more non-transitory computer-readable storage media of claim 16, wherein the plurality of compute nodes collectively issue only a desired number of de-allocation requests.

20. The one or more non-transitory computer-readable storage media of claim 16, wherein the estimated probability is determined based, in part, on a number of compute nodes within the plurality of compute nodes.