IP Library Granted Patent US 11,507,425
Granted Patent B2
US 11,507,425 · App. 16/951,157 · Granted Nov 22, 2022

Compute instance provisioning based on usage of physical and virtual components

Inventors: Chaoyi Huang (Shenzhen, CN); Jiongjiong Gu (Shenzhen, CN); Xiaoyong Min (Shenzhen, CN); Zhiyuan Cai (Shenzhen, CN)
Assignee: HUAWEI CLOUD COMPUTING TECHNOLOGIES CO., LTD.
G06F9/5027G06F9/4881G06F9/5088H04L67/1097H04L67/148H04L67/61
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 11,507,425
App. No.
16/951,157
Granted
Nov 22, 2022
Kind
B2
Abstract

According to a compute instance scheduling method, usage of a physical resource included in each computer, and usage of a physical resource occupied by a compute instance running on each computer in a compute instance provisioning system are collected. Then, a compute instance in the compute instance provisioning system is scheduled based on the usage of the physical resource in each computer and the usage of the physical resource occupied by the compute instance running on each computer.

Claims (79)

1. A method implemented by a compute instance provisioning system and comprising:

collecting a first usage of a first physical resource comprised in a first computer in the compute instance provisioning system, wherein the first computer is a physical component;

collecting a second usage of a second physical resource occupied by a first compute instance running on the first computer, wherein the first compute instance is a virtual component;

calculating a remaining physical resource parameter in the first computer based on the first usage;

calculating a first quality of service (QoS) parameter of the first compute instance based on the second usage;

calculating a second QoS parameter of the first computer based on the first QoS parameter; and

migrating the first compute instance or creating a second compute instance in the compute instance provisioning system based on the first usage, the second usage, the remaining physical resource parameter, and the second QoS parameter.

2. The method of claim 1 , further comprising:

obtaining a compute instance scheduling request instructing migration of the first compute instance or creation of the second compute instance;

selecting, based on the remaining physical resource parameter and the second QoS parameter, a second computer in the compute instance provisioning system that can deploy the first compute instance or the second compute instance; and

deploying the first compute instance or the second compute instance on the second computer.

3. The method of claim 1 , further comprising:

determining a second compute instance in the compute instance provisioning system whose third QoS parameter exceeds a QoS parameter threshold; and

performing QoS control on the second compute instance by:

isolating a third physical resource in a second computer on which the second compute instance is located for exclusive use by the second compute instance, or

scheduling a third compute instance running on the second computer.

4. The method of claim 3 , wherein the third physical resource comprises a memory device, a central processing unit (CPU) core, a last level cache (LLC), or a memory bandwidth.

5. The method of claim 3 , further comprising:

calculating, based on a fourth QoS parameter of a fourth compute instance running on the second computer, an interference coefficient of the fourth compute instance; and

scheduling, based on the interference coefficient, the fourth compute instance.

6. The method according to claim 3 , further comprising:

making, after performing the QoS control, a determination that the third QoS parameter still exceeds the QoS parameter threshold; and

migrating the second compute instance in response to the determination.

7. A computer program product comprising instructions that are stored on a computer-readable medium and that, when executed by a processor, cause a compute instance provisioning system to:

collect a first usage of a first physical resource comprised in a first computer in the compute instance provisioning system, wherein the first computer is a physical component;

collect a second usage of a second physical resource occupied by a first compute instance running on the first computer, wherein the first compute instance is a virtual component;

calculate a remaining physical resource parameter in the first computer based on the first usage;

calculate a first quality of service (QoS) parameter of the first compute instance based on the second usage;

calculate a second QoS parameter of the first computer based on the first QoS parameter; and

migrate the first compute instance or creating a second compute instance in the compute instance provisioning system based on the first usage, the second usage, the remaining physical resource parameter, and the second QoS parameter.

8. The computer program product of claim 7 , wherein the instructions further cause the compute instance provisioning system to:

obtain a compute instance scheduling request instructing migration of the first compute instance or creation of the second compute instance;

select, based on the remaining physical resource parameter and the second QoS parameter, a second computer in the compute instance provisioning system that can deploy the first compute instance or the second compute instance; and

deploy the first compute instance or the second compute instance on the second computer.

9. The computer program product of claim 7 , wherein the instructions further cause the compute instance provisioning system to:

determine a second compute instance in the compute instance provisioning system whose third QoS parameter exceeds a QoS parameter threshold; and

perform QoS control on the second compute instance by:

isolating a third physical resource in a second computer on which the second compute instance is located for exclusive use by the second compute instance, or

scheduling a third compute instance running on the second computer.

10. The computer program product of claim 9 , wherein the third physical resource comprises a memory device, a central processing unit (CPU) core, a last level cache (LLC), or a memory bandwidth.

11. The computer program product of claim 9 , wherein the instructions further cause the compute instance provisioning system to:

calculate, based on a fourth QoS parameter of a fourth compute instance running on the second computer, an interference coefficient of the fourth compute instance; and

schedule, based on the interference coefficient, the fourth compute instance.

12. The computer program product according to claim 9 , wherein the instructions further cause the compute instance provisioning system to:

make, after performing the QoS control, a determination that the third QoS parameter still exceeds the QoS parameter threshold; and

migrate the second compute instance in response to the determination.

13. A compute instance provisioning system comprising:

a memory configured to store instructions; and

a processor coupled to the memory and configured to execute the instructions to cause the compute instance provisioning system to:

collect a first usage of a first physical resource comprised in a first computer in the compute instance provisioning system, wherein the first computer is a physical component,

collect a second usage of a second physical resource occupied by a first compute instance running on the first computer, wherein the first compute instance is a virtual component,

calculate a remaining physical resource parameter in the first computer based on the first usage,

calculate a first quality of service (QoS) parameter of the first compute instance based on the second usage,

calculate a second QoS parameter of the first computer based on the first QoS parameter, and

migrate the first compute instance or creating a second compute instance in the compute instance provisioning system based on the first usage, the second usage, the remaining physical resource parameter, and the second QoS parameter.

14. The compute instance provisioning system of claim 13 , wherein the processor is further configured to execute the instructions to cause the compute instance provisioning system to:

obtain a compute instance scheduling request instructing migration of the first compute instance or creation of the second compute instance;

select, based on the remaining physical resource parameter and the second QoS parameter, a second computer in the compute instance provisioning system that can deploy the first compute instance or the second compute instance; and

deploy the first compute instance or the second compute instance on the second computer.

15. The compute instance provisioning system of claim 13 , wherein the processor is further configured to execute the instructions to cause the compute instance provisioning system to:

determine a second compute instance in the compute instance provisioning system whose third QoS parameter exceeds a QoS parameter threshold; and

perform QoS control on the second compute instance by:

isolating a third physical resource in a second computer on which the second compute instance is located for exclusive use by the second compute instance, or

scheduling a third compute instance running on the second computer.

16. The compute instance provisioning system of claim 15 , wherein the third physical resource comprises a memory device, a central processing unit (CPU) core, a last level cache (LLC), or a memory bandwidth.

17. The compute instance provisioning system of claim 15 , wherein the processor is further configured to execute the instructions to cause the compute instance provisioning system to:

calculate, based on a fourth QoS parameter of a fourth compute instance running on the second computer, an interference coefficient of the fourth compute instance; and

schedule, based on the interference coefficient, the fourth compute instance.

18. The method of claim 1 , wherein the physical component is an entity having a compute capability.

19. The method of claim 18 , wherein the entity is a personal computer (PC) or a server.

20. The method of claim 1 , wherein the virtual component is a computational unit configured to run a service.

21. The method of claim 20 , wherein the computational unit is a virtual machine (VM), a container, or a process.

22. The method of claim 1 , wherein migrating the first compute instance or creating the second compute instance comprises creating the second compute instance.

23. The method of claim 4 , wherein the third physical resource comprises the LLC.

24. The computer program product of claim 10 , wherein the third physical resource comprises the LLC.

25. The compute instance provisioning system of claim 15 , wherein the processor is further configured to execute the instructions to cause the compute instance provisioning system to:

make, after performing the QoS control, a determination that the third QoS parameter still exceeds the QoS parameter threshold; and

migrate the second compute instance in response to the determination.

26. The compute instance provisioning system of claim 16 , wherein the third physical resource comprises the LLC.

Assignments (2)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Mar 1, 2022
From: HUAWEI TECHNOLOGIES CO., LTD.
To: HUAWEI CLOUD COMPUTING TECHNOLOGIES CO., LTD.
Reel/Frame 059267/0088 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Dec 16, 2020
From: HUANG, CHAOYI; GU, JIONGJIONG; MIN, XIAOYONG; CAI, ZHIYUAN
To: HUAWEI TECHNOLOGIES CO., LTD.
Reel/Frame 054662/0732 →
Priority Claims (2)
CN 201911133776.X · Nov 19, 2019 · national
CN 201911424243.7 · Dec 31, 2019 · national
Continuity (1)
Related Publication 20210149735A1 · May 20, 2021