On-demand compute environment

Patent metadata

Field	Value
Publication number	US-12120040-B2
Application number	US-202217722076-A
Country	US
Kind code	B2
Filing date	Apr 15, 2022
Priority date	Mar 16, 2005
Publication date	Oct 15, 2024
Grant date	Oct 15, 2024

Abstract

An on-demand compute environment comprises a plurality of nodes within an on-demand compute environment available for provisioning and a slave management module operating on a dedicated node within the on-demand compute environment, wherein upon instructions from a master management module at a local compute environment, the slave management module modifies at least one node of the plurality of nodes.

First claim

What is claimed is:

1. A method of operating a compute environment comprising a plurality of resources so as to provide a plurality of logically independent clusters to respective ones of a plurality of users or organizations, the method comprising:
identifying a plurality of resource requirements associated with respective ones of the plurality of users or organizations;
causing partitioning of at least part of the plurality of resources based at least on the identified plurality of resource requirements, the partitioning enabling allocation of respective portions of the at least part of the plurality of resources to only the respective plurality of users or organizations for use thereby as a logically independent cluster such that each of the respective plurality of users or organizations have dedicated access to the respective portions of the at least part of the plurality of resources portioned therefor; and
presenting each of the plurality of users or organizations with respective access to only one or more aspects of the logically independent cluster allocated thereto, the presenting comprising (i) presenting only one or more first services which the respective user or organization utilizes or has utilized for processing of workload, and (ii) selectively excluding presentation of one or more second services which do not relate to the logically independent cluster of the respective user or organization.

2. The method of claim 1 , wherein:
the compute environment comprises one or more commonly managed clusters, each of the one or more clusters comprising a plurality of compute nodes; and
each of the plurality of users or organizations are separately or independently managed.

3. The method of claim 2 , further comprising:
receiving data indicative of one or more changes to at least one of the identified plurality of resource requirements; and
based at least on the received data, dynamically varying the allocation of at least one of the respective portions to accommodate the one or more changes.

4. The method of claim 3 , wherein the one or more changes comprise changes to one or more quality of service or service level requirements.

5. The method of claim 1 , wherein:
the identifying the plurality of resource requirements associated with each of the respective ones of the plurality of users or organizations comprises identifying at least one quality of service (QOS) or service level requirement associated with at least one of the plurality of users or organizations; and
the causing partitioning of at least part of the plurality of resources based at least on the identified plurality of resource requirements comprises causing partitioning so as to ensure the at least one QoS or service level requirement is at least met.

6. The method of claim 1 , wherein the presenting each of the plurality of users or organizations with respective access to only one or more aspects of the logically independent cluster allocated thereto, comprises presentation of only resources which the respective user or organization may utilize for processing of workload.

7. The method of claim 1 , wherein the presenting each of the plurality of users or organizations with respective access to only one or more aspects of the logically independent cluster allocated thereto, comprises presentation of only jobs or workload submitted by the respective user or organization.

8. The method of claim 1 , wherein the presenting each of the plurality of users or organizations with respective access to only one or more aspects of the logically independent cluster allocated thereto, comprises presentation of only one or more users or user credentials for the respective user or organization.

9. The method of claim 1 , wherein at least one of (i) the identifying a plurality of resource requirements associated with each of the respective ones of the plurality of users or organizations, or (ii) causing partitioning of at least part of the plurality of resources based at least on the identified plurality of resource requirements, comprises implementing at least one respective partitioning policy for each of the respective ones of the plurality of users or organizations.

10. The method of claim 1 , further comprising:
monitoring one or more conditions; and
based at least on the monitoring, dynamically varying the allocation of at least one of the respective portions.

11. The method of claim 10 , wherein the dynamically varying the allocation of at least one of the respective portions comprises dynamically varying in accordance with one or more virtual cluster-specific or user-specific policies.

12. The method of claim 10 , wherein the dynamically varying the allocation of at least one of the respective portions comprises dynamically varying in accordance with one or more policies applicable to an entirety of the compute environment.

13. The method of claim 10 , wherein the one or more conditions comprise a then-current load on at least one of (i) the compute environment, or (ii) the one or more logically independent clusters.

14. The method of claim 10 , wherein the one or more conditions comprises a temporal condition or state.

15. The method of claim 14 , wherein the temporal condition or state comprises one of (i) a guaranteed or required response time for processing of workload, or (ii) a particular calendared event.

16. A method of providing virtual compute clusters to respective ones of a plurality of users using one or more commonly managed compute environments, the method comprising:
identifying a plurality of resource requirements associated with respective ones of the plurality of users; and
causing partitioning of resources of the one or more commonly managed compute environments based at least on the identified plurality of resource requirements thereby generating respective partitions, the partitioning enabling allocation of respective portions of the resources to the respective plurality of users for concurrent use thereby as respective virtual compute clusters, wherein each of the respective plurality of users have exclusive access to the respective portions of the resources partitioned therefor; and
wherein the partitioning is performed so as to (i) at least provide at least some of the plurality of users with guaranteed availability of resources allocated to their respective virtual compute cluster for at least a period of time, and (ii) allow modification of the resources in support of one or more policies affecting both a) the at least some of the plurality of users and b) other users not part of the plurality of users, the modification being transparent to the at least some of the plurality of users.

17. The method of claim 16 , wherein the sharing of at least some of the resources by two or more of the plurality of users comprises sharing which is dynamically modified as a function of at least one of a) time, or b) load on at least one of the virtual compute clusters.

18. The method of claim 16 , further comprising providing load-balancing between at least two of the virtual compute clusters, the providing load-balancing between at least two of the virtual compute clusters comprises providing the load-balancing between at least two of the at least some of the plurality of users having guaranteed resource availability.

19. The method of claim 16 , wherein the guaranteed resource availability is sufficient to at least meet the plurality of resource requirements associated with the at least some of the plurality of users.

20. The method of

Assignees

• Iii Holdings 12 Llc

Inventors

• Jackson David Brian

Classifications

• H04L47/80 Primary

  Actions related to the user profile or the type of traffic · CPC title

• H04L47/83

  based on usage prediction · CPC title

• G06F9/5027 Primary

  the resource being a machine, e.g. CPUs, Servers, Terminals · CPC title

• H04L67/125

  involving control of end-device applications over a network · CPC title

• H04L67/10

  in which an application is distributed across nodes in the network (software deployment G06F8/60; multiprogramming arrangements G06F9/46) · CPC title