CPU SCHEDULER CONFIGURED TO SUPPORT LATENCY SENSITIVE VIRTUAL MACHINES

• US 20150058861A1
• Filed: 08/25/2014
• Published: 02/26/2015
• Est. Priority Date: 08/26/2013
• Status: Active Grant

• Current Assignee: Vmware LLC (Broadcom, Inc.)
• Original Assignee: VMware, Inc. (Broadcom, Inc.)
• Inventors: ZHENG, Haoqiang, SINGARAVELU, Lenin, AGARWAL, Shilpi, HECHT, Daniel Michael, SMITH, Garrett
• Granted Patent: US 9,262,198 B2
• Time in Patent Office: Days
• US Class Current: 718/104
• CPC Class Codes: G06F 13/24 using interrupt G06F13/32 takes precedence
• CPC Class Codes: G06F 2009/4557 Distribution of virtual machine instances; Migration and load balancing
• CPC Class Codes: G06F 2009/45575 Starting, stopping, suspending or resuming virtual machine instances
• CPC Class Codes: G06F 2009/45591 Monitoring or debugging support
• CPC Class Codes: G06F 2209/5017 Task decomposition
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/4812 by interrupt, e.g. masked
• CPC Class Codes: G06F 9/4887 involving deadlines, e.g. rate based, periodic
• CPC Class Codes: G06F 9/50 Allocation of resources, e.g. of the central processing unit [CPU]
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals
• CPC Class Codes: G06F 9/5033 considering data affinity
• CPC Class Codes: G06F 9/5088 involving task migration
• CPC Class Codes: H04L 43/0894 Packet rate
• CPC Class Codes: H04L 43/16 Threshold monitoring
• CPC Class Codes: H04L 47/56 implementing delay-aware scheduling
• CPC Class Codes: H04L 47/801 Real time traffic

Abstract

A host computer has one or more physical central processing units (CPUs) that support the execution of a plurality of containers, where the containers each include one or more processes. Each process of a container is assigned to execute exclusively on a corresponding physical CPU when the corresponding container is determined to be latency sensitive. The assignment of a process to execute exclusively on a corresponding physical CPU includes the migration of tasks from the corresponding physical CPU to one or more other physical CPUs of the host system, and the directing of task and interrupt processing to the one or more other physical CPUs. Tasks of the process corresponding to the container are then executed on the corresponding physical CPU.

Claims

1. In a host computer having one or more physical central processing units (CPUs) that support the execution of a plurality of containers, the containers each including one or more processes, wherein the host computer is programmed to perform a method of assigning at least one of the processes to have exclusive affinity to a corresponding physical CPU, the method comprising:

determining that a first container is latency sensitive; and

responsive to the determining, assigning each of the processes of the first container to have exclusive affinity to one or more corresponding physical CPUs, said assigning comprising;

migrating running tasks on the corresponding physical CPUs to the one or more other physical CPUs of the host system;

directing queued tasks and interrupt processing for the corresponding physical CPUs to the one or more other physical CPUs; and

executing tasks of each of the processes of the first container on the one or more corresponding CPUs to which the process has exclusive affinity.

2. A system, comprising:

a host computer having a memory and one or more physical central processing units (CPUs) that support the execution of a plurality of containers, the containers each including one or more processes, wherein the host computer is programmed to perform a method of assigning at least one of the processes to have exclusive affinity to a corresponding physical CPU, the method comprising;

determining that a first container is latency sensitive; and

responsive to the determining, assigning each of the processes of the first container to have exclusive affinity to one or more corresponding physical CPUs, said assigning comprising;

migrating running tasks on the corresponding physical CPUs to the one or more other physical CPUs of the host system;

directing queued tasks and interrupt processing for the corresponding physical CPUs to the one or more other physical CPUs; and

executing tasks of each of the processes of the first container on the one or more corresponding CPUs to which the process has exclusive affinity.

3. The system of claim 2, wherein the first container is a first virtual machine and each of the one or more processes included in the first container is a virtual CPU, and at least one virtual CPU is assigned to have exclusive affinity to a corresponding physical CPU.

4. The system of claim 3, the method further comprising:

detecting a change to a state of the first virtual machine, wherein detecting the change to the state of the first virtual machine comprises detecting that the state of the first virtual machine has been changed from powered-off to powered-on.

5. The system of claim 4, wherein detecting the change to the state of the first virtual machine comprises detecting that a latency sensitivity indicator for the first virtual machine has been changed.

6. The system of claim 3, the method further comprising, prior to the assigning of the virtual CPUs, determining that the host computer has a sufficient number of physical CPUs to which the virtual CPUs are to be assigned.

7. The system of claim 6, wherein the method further comprises reserving a predetermined number of physical CPUs of the host computer to process tasks that do not correspond to the first virtual machine.

8. The system of claim 3, wherein determining that the first virtual machine is latency sensitive comprises:

reading a latency sensitivity indicator for the first virtual machine;

reading a CPU entitlement value for the first virtual machine;

determining that the latency sensitivity indicator is a predetermined value; and

determining that the CPU entitlement value is a maximum value.

9. The system of claim 8, wherein the maximum value for the CPU entitlement value indicates that each virtual CPU of the first virtual machine is permitted to consume the full processing capacity of one of the physical CPUs.

10. The system of claim 3, wherein migrating the processes from the corresponding physical CPUs comprises:

halting a task that is currently being executed by one of the corresponding physical CPUs; and

placing the halted task into a run queue of one of the one or more other physical CPUs.

11. The system of claim 10, wherein migrating the processes from the corresponding physical CPUs further comprises:

copying a task from a run queue of one of the corresponding physical CPUs to a run queue of one of the one or more other physical CPUs; and

removing the task from the run queue of the corresponding physical CPU.

12. The system of claim 3, wherein the method further comprises disabling frequency scaling for each of the corresponding physical CPUs.

13. A non-transitory computer-readable medium comprising instructions executable by a host computer, the host computer having one or more physical central processing units (CPUs) that support the execution of a plurality of containers, the containers each including one or more processes, where the instructions, when executed, cause the host computer to perform a method of assigning processes to execute exclusively on a corresponding physical CPU, the method comprising:

determining that a first container is latency sensitive; and

responsive to the determining, assigning each of the processes of the first container to have exclusive affinity to one or more corresponding physical CPUs, said assigning comprising;

migrating running tasks from the corresponding physical CPUs to the one or more other physical CPUs of the host system;

directing queued tasks and interrupt processing for the corresponding physical CPUs to the one or more other physical CPUs; and

executing tasks of each of the processes of the first container on the one or more corresponding CPUs to which the process has exclusive affinity.

14. The computer-readable medium of claim 13, wherein the first container is a first virtual machine and each of the one or more processes included in the first container is a virtual CPU, and at least one virtual CPU is assigned to have exclusive affinity to a corresponding physical CPU.

15. The computer-readable medium of claim 14, the method further comprising:

detecting a change to a state of the first virtual machine, wherein detecting the change to the state of the first virtual machine comprises detecting that the state of the first virtual machine has been changed from powered-off to powered-on.

16. The computer-readable medium of claim 15, wherein detecting the change to the state of the first virtual machine further comprises detecting that a latency sensitivity indicator for the first virtual machine has been changed.

17. The computer-readable medium of claim 14, wherein the method further comprises, prior to the assigning of the virtual CPUs, determining that the host computer has a sufficient number of physical CPUs to which the virtual CPUs are to be assigned.

18. The computer-readable medium of claim 17, wherein the method further comprises reserving a predetermined number of physical CPUs of the host computer to process tasks that do not correspond to a latency sensitive virtual machine.

19. The computer-readable medium of claim 14, wherein determining that the first virtual machine is latency sensitive comprises:

reading a latency sensitivity indicator for the first virtual machine;

reading a CPU entitlement value for the first virtual machine

determining that the latency sensitivity indicator is a predetermined value; and

determining that the CPU entitlement value is a maximum value.

20. The computer-readable medium of claim 19, wherein the maximum value for the CPU entitlement value indicates that each virtual CPU of the first virtual machine is permitted to consume the full processing capacity of one of the physical CPUs.