Resource sharing in computer clusters according to objectives

• US 8,972,579 B2
• Filed: 09/06/2012
• Issued: 03/03/2015
• Est. Priority Date: 09/06/2012
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Simitsis, Alkiviadis, Palpanas, Themistoklis, Harizopoulos, Stavros, Iori, Enrico, Wilkinson, William K
• Primary Examiner(s): HIGA, BRENDAN Y
• Application Number:
  US13/605,651
• Publication Number:
  US 20140068055A1
• Time in Patent Office: 908 Days
• Field of Search: 709223-226, 709/229
• US Class Current: 709/226
• CPC Class Codes: G06F 2209/5014 Reservation
• CPC Class Codes: G06F 2209/508 Monitor
• CPC Class Codes: G06F 9/5011 the resources being hardware resources other than CPUs, Servers and Terminals
• CPC Class Codes: G06F 9/5094 where the allocation takes into account power or heat criteria power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

A method of assigning resources of a computer duster with resource sharing according to objectives. The method includes monitoring resources of each of a plurality of cloud nodes, providing information descriptive of the cloud node resources, receiving a reservation, determining whether resources are available to satisfy the reservation and any other pending reservations, if resources are available, using a rapid search to determine resource assignments for the reservation and any other pending reservations according to one or more objectives, and allocating resources according to the resource assignments.

Claims

1. A method of assigning resources of a computer cluster with resource sharing according to objectives, the method comprising:

monitoring resources of each of a plurality of cloud nodes;

providing information descriptive of the cloud node resources;

receiving a reservation;

determining whether resources are available to satisfy the reservation and any other pending reservations;

if resources are available, using a rapid search to determine resource assignments for the reservation and any other pending reservations according to one or more objectives; and

allocating resources according to the resource assignments.

2. The method of claim 1 wherein the one or more objectives comprise minimizing power consumption throughout the computer cluster.

3. The method of claim 2 wherein the rapid search comprises a rigid-resource search that assigns cloud nodes for reservations in increasing order of power consumed per core.

4. The method of claim 2 wherein the rapid search comprises a flexible-resource search that assigns cloud nodes for reservations in increasing order of power consumed per core.

5. The method of claim 4 wherein the flexible-resource search comprises, for a reservation to which resources from more than one cloud node are assigned, apportioning the resources from each cloud node in relative quantities determined by relative quantities of resources specified in the reservation.

6. The method of claim 2 wherein the rapid search comprises a flexible-start search that assigns cloud nodes for reservations in increasing order of power consumed per core.

7. The method of claim 6 wherein the flexible-start search comprises assigning cloud nodes for reservations in order of start times.

8. The method of claim 6 wherein the flexible-start search comprises assigning cloud nodes for reservations having fixed start times and then assigning cloud nodes for reservations having flexible start times according to intervals of time during which the cloud nodes have sufficient resources to accommodate the flexible-start reservations.

9. The method of claim 1 wherein determining whether resources are available comprises making the determination for all sub-tasks of a composite reservation.

10. A computer cluster with resource sharing according to objectives, the cluster comprising a plurality of cloud nodes each including one or more resources;

a terminal;

data storage; and

an allocation node to;

monitor the cloud node resources;

provide information descriptive of the cloud node resources to a customer through the terminal;

receive a reservation for cloud node resources from the customer;

store the reservation in the data storage;

determine assignments of the cloud node resources for the reservation and any other pending reservations according to one or more objectives by rapid search; and

allocate the cloud node resources to customers according to the resource assignments.

11. The computer cluster of claim 10 wherein the one or more objectives comprise minimizing power consumption throughout the computer cluster.

12. The computer cluster of claim 11 wherein the rapid search comprises a rigid-resource search to assign cloud nodes in increasing order of power consumed per core.

13. The computer cluster of claim 11 wherein the rapid search comprises a flexible-resource search to assign cloud nodes in increasing order of power consumed per core.

14. The computer cluster of claim 11 wherein the rapid search comprises a flexible-start search to assign cloud nodes in increasing order of power consumed per core.