Service level based priority scheduler for multi-tenancy computing systems

• US 10,462,070 B1
• Filed: 06/30/2016
• Issued: 10/29/2019
• Est. Priority Date: 06/30/2016
• Status: Active Grant

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Inventors: Smaldone, Stephen, Wigmore, Ian, Shvaiger, Felix, Don, Arieh, Benhanokh, Gabi
• Primary Examiner(s): Springer, James E
• Application Number:
  US15/198,284
• Time in Patent Office: 1,216 Days
• Field of Search: 709226
• CPC Class Codes: H04L 41/22 comprising specially adapted graphical user interfaces [GUI]
• CPC Class Codes: H04L 41/5003 Managing SLA; Interaction between SLA and QoS
• CPC Class Codes: H04L 41/5022 by giving priorities, e.g. assigning classes of service
• CPC Class Codes: H04L 47/2433 Allocation of priorities to traffic types
• CPC Class Codes: H04L 47/2458 Modification of priorities while in transit
• CPC Class Codes: H04L 47/805 QOS or priority aware
• CPC Class Codes: H04L 47/821 Prioritising resource allocation or reservation requests

Abstract

Systems and methods for scheduling multiple flows in a computing system or for allocating resources to the flows in the computing system. Each flow may be associated with a target priority and a target performance. A current priority can be determined for each flow and resources are allocated based on the highest current priority. Over time, the current priorities change and each flow will receive resources as the flows are scheduled for execution. No flow is starved of resources.

Claims

1. A method for allocating resources in a computing system providing a computing service to one or more clients, the method comprising:

receiving parameters from the one or more clients for multiple flows associated with the one or more clients, the parameters including at least a target priority and a target performance for each of the multiple flows;

for each of the multiple flows, converting the target performance into a measurable characteristic and determining a current performance based on the measurable characteristic;

determining a current priority for each of the multiple flows executing in the computing system, wherein the current priority is based on the target priority, the target performance and the current performance;

scheduling the multiple flows for a schedule block based on the current priorities, wherein each schedule block corresponds to a predetermined period of time;

allocating resources to the multiple flows based on relative current priorities of the multiple flows for the schedule block;

updating the current priority of each of the multiple flows for a next schedule block; and

reallocating the resources to the multiple flows for the next schedule block based on the relative updated current priorities.

2. The method of claim 1, wherein the target priority comprises a number and wherein the target performance comprises a time period or a rate.

3. The method of claim 1, further comprising determining the current priority for each flow by:

multiplying the target priority of the flow by a ratio of the target performance to the current performance.

4. The method of claim 1, wherein the one or more parameters of each of the multiple flows further include an excess priority and a second priority, wherein the second priority is associated with achieving the target performance.

5. The method of claim 4, wherein, when the current performance is greater than or equal to a target performance, the current priority of a flow is equal to the excess priority multiplied by a ratio of the target performance to the current performance.

6. A method of scheduling flows in a computing environment, the method comprising:

providing storage to multiple tenants in a computing system configured to provide a computing service to the tenants, wherein the storage of each tenant is associated with a storage quota, wherein each entity is associated with multiple objects and each object is associated with a flow;

receiving parameters from the tenants in for the flows associated with the tenants, the parameters including at least a target priority and a target performance for each of the flows;

determining a current performance for each of the flows by converting the target performance into a measurable characteristic;

determining a current priority for each flow executing in the computing system based on the target performance;

scheduling the flows for a schedule block based on the current priorities, wherein each schedule block corresponds to a predetermined period of time and wherein the flow with the highest priority is given priority access to resources of the computing system for the schedule block;

for a next schedule block, updating the current priorities of the flows and giving prioritized access to the resources to the flow with the highest updated current priority during the next schedule block.

7. The method of claim 6, further comprising determining the current priority, when a current performance is less than a target performance by multiplying a target priority by a ratio of a target performance to a current performance.

8. The method of claim 7, further comprising determining, when the current performance is greater than or equal to the target performance, by multiplying an excess priority by the ratio of the target performance to the current performance.

9. The method of claim 6, wherein the computing service include backup services.

10. The method of claim 9, wherein each flow comprises a process associated with a backup operation.

11. The method of claim 6, further comprising receiving parameters through a user interface, the parameters including a target priority, a target performance, and an excess priority.

12. The method of claim 6, wherein the current priority of a flow increases the further away the current performance is from the target performance.

13. The method of claim 6, wherein the resources include processor resources.

14. A computing system configured to provide a computing service to one or more clients, wherein each client is associated with one or more objects, the computing system comprising:

a storage device, wherein the objects are stored in the storage;

computing resources including at least processing resources; and

a scheduler that includes computer executable instructions that, when executed, perform a method for scheduling flows associated with the objects, the method comprising;

receiving parameters from the one or more clients for multiple flows associated with the one or more clients, the parameters including at least a target priority and a target performance for each of the multiple flows;

for each of the multiple flows, determining a current performance by converting the target performance into a measurable characteristic;

determining a current priority for each of the flows executing in the computing system, wherein the current priority is based on the target priority, the target performance and a current performance when the current performance is less than the target performance and

wherein the current priority is based on an excess priority, the target performance and the current performance when the current performance is equal to or greater than the target performance;

scheduling the multiple flows for a schedule block based on the current priorities, wherein each schedule block corresponds to a predetermined period of time;

allocating resources to the multiple flows based on relative current priorities of the multiple flows for the schedule block;

updating the current priority of each of the multiple flows for a next schedule block; and

reallocating the resources to the multiple flows for the next schedule block based on the relative updated current priorities.

15. The computing system of claim 14, further comprising presenting a graph to a user to aid in setting the target priority and the target performance.