Self-organized retail source request routing and distributed load sharing systems and methods

• US 10,673,936 B2
• Filed: 12/29/2017
• Issued: 06/02/2020
• Est. Priority Date: 12/30/2016
• Status: Active Grant

• Current Assignee: Walmart Apollo, LLC (WalMart Inc.)
• Original Assignee: Walmart Apollo, LLC (WalMart Inc.)
• Inventors: Maddali, Praveen Kumar, Bansal, Parijat, Ranganath, Prithvi
• Primary Examiner(s): Gillis, Brian J.
• Assistant Examiner(s): Ling, Chhian (Amy)
• Application Number:
  US15/858,388
• Publication Number:
  US 20180191818A1
• Time in Patent Office: 886 Days
• CPC Class Codes: H04L 47/125 by balancing the load, e.g. traffic engineering
• CPC Class Codes: H04L 47/765 triggered by the end-points
• CPC Class Codes: H04L 47/803 Application aware
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/1014 based on the content of a request
• CPC Class Codes: H04L 67/1025 Dynamic adaptation of the criteria on which the server selection is based
• CPC Class Codes: H04L 67/566 Grouping or aggregating service requests, e.g. for unified processing
• CPC Class Codes: H04L 67/63 Routing a service request depending on the request content or context

Abstract

Some embodiments provide self-organizing retail resource load sharing systems comprising: a cluster of a set of node systems that receive resource requests; multiple different resource systems that implement predefined tasks consistent with the resource requests; wherein the node systems: access a valid ownership allocation configuration; confirm when a first node system is currently defined as the owner of a first resource system and authorizing activation of the corresponding first resource system when the resource request is intended for the first resource system; and cause the first node system to operate as a proxy and communicate the first resource request to a different embedded allocation application on a second node system that currently owns a second resource system when the first resource request is intended for the second resource system. Typically, the self-organizing load sharing is provided without external coordinators and reduced infrastructure, while providing high resource availability and reliability.

Claims

1. A retail resource load sharing system, comprising:

a cluster of a set of multiple node systems communicatively coupled with a separate load balancer and configured to receive resource requests from the load balancer distributing the resource requests across the multiple node systems;

multiple different resource systems configured to each implement at least one predefined task of multiple different predefined tasks consistent with the resource requests, wherein each of the different resource systems is communicatively accessible over a communication network;

wherein each of the multiple node systems comprises a node control circuit and an embedded allocation application configured to communicate with the other embedded allocation applications of the set of multiple node systems, wherein the multiple embedded allocation applications are configured to apply resource allocation rules from the first set of resource allocation rules to cooperatively define a resource ownership allocation configuration that dictates ownership of each of the resource systems exclusively to one of the multiple node systems such that an owner node system exclusively has access to implement the corresponding task performed by the resource system while the ownership allocation configuration is valid and accessible across the multiple node systems;

wherein each of the node system control circuits is further configured to;

access the valid ownership allocation configuration in response to receiving a first resource request from the load balancer;

confirm, based on the ownership allocation configuration, when a first node system is currently defined as the owner of a first resource system and authorizing activation of the corresponding first resource system of the multiple resource systems when the first resource request is intended for the first resource system;

cause the first node system to operate, based on an ownership allocation rule of the first set of resource allocation rules, as a proxy and communicate the first resource request to a different embedded allocation application on a second node system that currently owns a second resource system in accordance with the ownership allocation configuration when the first resource request is intended for the second resource system; and

wherein the embedded allocation applications cooperatively operate to implement a load sharing system and cooperatively operate to maintain and update a cache, maintained across the multiple node systems, storing ownership allocation configuration that is currently valid at each of the node systems.

2. The system of claim 1, wherein at least a first embedded allocation application is configured to apply one or more resource allocation rules of the first set of resource allocation rules, in response to receiving a notification communicated to the embedded allocation applications of the multiple node systems by a third embedded allocation application on a third node system initially joining the cluster of the set of multiple node systems, to reallocate the resource systems and define a revised resource ownership allocation configuration that dictates ownership of each of the resource systems to change ownership of a plurality of the resource systems, which were defined to be owned by one or more of the set of node systems, to be defined as exclusively owned by the third node system.

3. The system of claim 2, wherein the second node system is configured to disable control over a third resource system based on the revised ownership allocation configuration;

and the third node system is configured to communicatively establish control over the third resource system based on the revised ownership allocation configuration.

4. The system of claim 2, wherein the first embedded allocation application in applying the resource allocation rules of the first set of resource allocation rules evaluates task loads of each of the multiple resource systems, and defines the revised resource ownership allocation configuration based on the task loads of the multiple resource systems.

5. The system of claim 1, wherein at least a first embedded allocation application is configured to detect the second node system is no longer functioning as part of the cluster, identify a subset of the resource systems comprising each of the resource systems owned by the second node system as dictated by the ownership allocation configuration, apply one or more resource allocation rules of the first set of resource allocation rules to reallocate each of the subset of resource systems and define a revised resource ownership allocation configuration that dictates ownership of each of the subset of resource systems to change ownership of each of the subset of the resource systems to be defined as exclusively owned by one of the remaining node systems of the cluster.

6. The system of claim 1, wherein at least a first embedded allocation application implemented on the first node system is configured to identify that a second embedded allocation application implemented on the second node system was defined as a master embedded allocation application, and takes over as the master embedded allocation application.

7. The system of claim 1, wherein the first node system of the cluster is configured to:

identify that a second resource request, issued by the load balancer, is received by the first node system wherein the second resource request is directed to be implemented by a first resource system;

access the resource ownership allocation configuration;

determine based on the resource ownership allocation configuration that the first node system is currently defined as the owner of the first resource system; and

initiate a communication from the first node system to the first resource system to cause the activation of the first resource system to implement the second resource request.

8. The system of claim 1, wherein a first node system in operating as the proxy is configured to:

identify that a third resource request, issued by the load balancer, is received by the first node system wherein the third resource request is directed to be implemented by the second resource system;

access, through a first embedded allocation application implemented on the first node system, the resource ownership allocation configuration;

determine based on the resource ownership allocation configuration that the second node system is currently defined as the owner of the second resource system; and

initiate a communication to a second embedded allocation application implemented through the second node system directing the second embedded allocation application to initiate performance of the third resource request.

9. The system of claim 8, wherein the second embedded allocation application on the second node system is configured to identify that a response is received from the second resource system in response to the third resource request, and to cause the response to be relayed to the first node system;

and

wherein the first embedded allocation application is configured to cause the first node system to communicate the relayed response to be directed to a requesting client system that issued the third resource request.

10. The system of claim 1, wherein the embedded allocation applications are further configured to form socket connections between two or more node systems and enables data transfers between two or more of the embedded allocation applications utilized in defining the ownership allocation configuration.

11. The system of claim 10, wherein each node system of the set of multiple node systems comprises cache memory storing parameters and allocation data, and wherein the embedded allocation applications of the multiple node systems provide distributed cache capabilities that is accessible across and to each of the multiple node systems.

12. A method of load sharing retail resource systems, comprising:

by each of a plurality of node systems, each implementing a respective embedded allocation application, of a cluster comprising a set of multiple node systems that are communicatively coupled with a separate load balancer and configured to receive resource requests from the load balancer distributing the resource requests across the multiple node systems;

accessing a valid ownership allocation configuration in response to receiving a first resource request from the load balancer, wherein the resource ownership allocation configuration is cooperatively defined by the multiple embedded allocation applications and dictates exclusive ownership of each resource system, of multiple resource systems, to one of the multiple node systems such that an owner node system exclusively has access to implement a corresponding task performed by the owned resource system while the ownership allocation configuration is valid and accessible across the multiple node systems;

confirming, based on the ownership allocation configuration, when a first node system is currently defined as the owner of a first resource system and authorizing activation of the corresponding first resource system of the multiple resource systems currently owned by the correspond first node system when the first resource request is intended for the first resource system;

causing the first node system to operate, based on an ownership allocation rule of the first set of resource allocation rules, as a proxy and communicate the first resource request to a different embedded allocation application on a second node system that currently owns a second resource system in accordance with the ownership allocation configuration when the first resource request is intended for the second resource system; and

wherein the embedded allocation applications cooperatively operate to implement a load sharing system and cooperatively operate to maintain and update a cache, maintained across the multiple node systems, storing ownership allocation configuration that is currently valid at each of the node systems.

13. The method of claim 12, further comprising:

applying, by a first embedded allocation application of the plurality of embedded allocation applications, one or more resource allocation rules of the first set of resource allocation rules, in response to receiving a notification communicated to the embedded allocation applications of the multiple node systems by a third embedded allocation application on a third node system initially joining the cluster of the set of multiple node systems, to reallocate the resource systems; and

defining a revised resource ownership allocation configuration based on the reallocation that dictates ownership of each of the resource systems to change ownership of a plurality of the resource systems, which were defined to be owned by one or more of the set of node systems, to be defined as exclusively owned by the third node system.

14. The method of claim 13, further comprising:

causing the second node system to disable control over a third resource system based on the revised ownership allocation configuration; and

causing the third node system to communicatively establish control over the third resource system based on the revised ownership allocation configuration.

15. The method of claim 13, wherein the applying one or more resource allocation rules of the first set of resource allocation rules comprises evaluating task loads of each of the multiple resource systems;

and

wherein the defining the revised resource ownership allocation configuration comprises defining the revised resource ownership allocation configuration based on the task loads of the multiple resource systems.

16. The method of claim 12, further comprising:

detecting, by a first embedded allocation application of the plurality of embedded allocation applications, that the second node system is no longer functioning as part of the cluster;

identifying a subset of the resource systems comprising each of the resource systems owned by the second node system as dictated by the ownership allocation configuration;

applying one or more resource allocation rules of the first set of resource allocation rules and reallocating each of the subset of resource systems; and

defining a revised resource ownership allocation configuration that dictates ownership of each of the subset of resource systems to change ownership of each of the subset of the resource systems to be defined as exclusively owned by one of the remaining node systems of the cluster.

17. The method of claim 12, further comprising:

identifying, by a first embedded allocation application on the first node system of the cluster, that a second resource request, issued by the load balancer, is received by the first node system wherein the second resource request is directed to be implemented by a first resource system;

accessing the resource ownership allocation configuration;

determining based on the resource ownership allocation configuration that the first node system is currently defined as the owner of the first resource system; and

initiating a communication from the first node system to the first resource system to cause the activation of the first resource system to implement the second resource request.

18. The method of claim 12, wherein the first embedded allocation application in operating as the proxy is configured to:

identifying, by the first embedded allocation application of the plurality of embedded allocation applications, that a third resource request, issued by the load balancer, is received by the first node system wherein the third resource request is directed to be implemented by the second resource system;

accessing the resource ownership allocation configuration;

determining based on the resource ownership allocation configuration that the second node system is currently defined as the owner of the second resource system; and

initiating a communication to a second embedded allocation application implemented through the second node system directing the second embedded allocation application to initiate performance of the third resource request.

19. The method of claim 18, further comprising:

identifying, by the second embedded allocation application, that a response is received from the second resource system in response to the third resource request, and causing the response to be relayed to the first node system; and

causing the first node system to communicate the relayed response to be directed to a requesting client system that issued the third resource request.