VIRTUAL WORKLOAD MIGRATIONS

• US 20200133702A1
• Filed: 10/30/2018
• Published: 04/30/2020
• Est. Priority Date: 10/30/2018
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Inventors: Sharma, Ranjeet Kumar, Kadambala, Anusha, Juturu, Raja Sekhar Reddy
• Granted Patent: US 11,150,931 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 2009/4557 Distribution of virtual machine instances; Migration and load balancing
• CPC Class Codes: G06F 2009/45591 Monitoring or debugging support
• CPC Class Codes: G06F 2209/5019 Workload prediction
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/4856 resumption being on a different machine, e.g. task migration, virtual machine migration G06F9/5088 takes precedence
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: G06F 9/5088 involving task migration

Abstract

An example system including a plurality of computing resources, distributed across a plurality of hosts, to execute virtual workloads; and a computing device, communicatively coupled to the plurality of hosts, comprising a processing resource and a memory resource. The memory resource may store instructions executable by the processing resource to monitor utilization data of the plurality of computing resources executing assigned respective virtual workloads; predict a destination host from the plurality of hosts with available computing resources in an amount to accommodate a predicted resource utilization of a particular virtual workload based on the monitored utilization data of the plurality of computing resources; and migrate the particular virtual workload assigned to a source host of the plurality of hosts to the destination host to be executed.

Claims

1. A system comprising:

a plurality of computing resources, distributed across a plurality of hosts, to execute virtual workloads; and

a computing device, communicatively coupled to the plurality of hosts, comprising a processing resource and a memory resource to store instructions executable by the processing resource to;

monitor utilization data of the plurality of computing resources executing assigned respective virtual workloads;

predict a destination host from the plurality of hosts with available computing resources in an amount to accommodate a predicted resource utilization of a particular virtual workload based on the monitored utilization data of the plurality of computing resources; and

migrate the particular virtual workload assigned to a source host of the plurality of hosts to the destination host to be executed.

2. The system of claim 1, wherein the plurality of computing resources is distributed across a hybrid cloud.

3. The system of claim 1, wherein the utilization data of the plurality of computing resources includes a CPU utilization metric determined at each host of the plurality of hosts.

4. The system of claim 1, wherein the utilization data of the plurality computing resources includes a memory utilization metric determined at each host of the plurality of hosts.

5. The system of claim 1, wherein the utilization data of the plurality computing resources includes an amount of containers operating and the computational load imposed by the amount of containers operating at each host of the plurality of hosts.

6. The system of claim 1, wherein the computing device includes instructions executable by the processing resource to predict a computational load at the source host when the particular virtual workload is to be executed and to trigger the instructions to migrate based on the computational load exceeding a threshold.

7. The system of claim 1, wherein the computing device includes instructions executable by the processing resource to predict the available computing resources based on a computational load predicted to be imposed upon the destination host by other virtual workloads when the particular workload is to be executed.

8. The system of claim 1, wherein the utilization data of the plurality of computing resources includes a network bandwidth utilization metric determined at each host of the plurality of hosts.

9. A non-transitory machine-readable storage medium having stored thereon machine-readable instructions to cause a computer processor to:

collect resource utilization data from each host device of a plurality of host devices executing respective assigned virtual workloads and from a particular virtual workload;

identify a utilization pattern for a destination host device of the plurality of host devices and a utilization pattern for the particular virtual workload based on the utilization data; and

create migration parameters to migrate the particular virtual workload from a source host device of the plurality of host devices to the destination host device based on the identified utilization pattern for the destination host device and the identified utilization pattern for the particular virtual workload.

10. The medium of claim 9, wherein the instructions to identify the utilization pattern for the destination host device based on the utilization data includes instructions to identify the utilization patterns from historical utilization data and live utilization data.

11. The medium of claim 9, wherein the utilization pattern for the destination host device is identified based on utilization data corresponding to a geo-location of the destination host device.

12. The medium of claim 9, wherein the utilization pattern for the destination host device is identified based on utilization data corresponding to a period of time that the workload is to be executed by the destination host device.

13. The medium of claim 9, including instructions to:

predict, from the utilization pattern for the particular virtual workload, a resource utilization imposed to execute the particular virtual workload;

predict, from the utilization pattern for the source host device, an availability of computing resources and network resources at the source host device to execute the particular virtual workload.

14. The medium of claim 13, including instructions to:

predict, based on the utilization pattern for the destination host device, an amount of computing resources and network resources available at the destination host device to execute the particular virtual workload; and

determine whether the destination host device is compatible with the particular virtual workload based on the predicted resource utilization to execute the particular virtual workload and the predicted amount of computing resources and network resources available at the destination host device when the particular virtual workload is to be executed.

15. The medium of claim 9, wherein the destination host device is in a different data center than the source host device.

16. A method, comprising:

monitoring computing resource utilization and network resource utilization by each host of a plurality of hosts executing assigned virtual workloads;

predicting, based on the monitored computing resource utilization and the monitored network resource utilization, a utilization load on a source host of the plurality of hosts imposed by an execution of a workload assigned to the source host;

predicting, based on the monitored computing resource utilization and the monitored network resource utilization, available computational capacity and network capacity at a destination host of the plurality of hosts; and

migrating the virtual workload assigned to the source host to the destination host to be executed by the available computational capacity and network capacity at the destination host.

17. The method of claim 16, including executing a second workload assigned to the destination host at the destination host when simultaneously executing the virtual workload migrated to the destination host from the source host.

18. The method of claim 16, including determining whether the available computational capacity and network capacity at the destination host accommodates an execution of the virtual workload.

19. The method of claim 16, including determining a carbon footprint associated with an execution of the virtual workload at the source host and a carbon footprint associated with an execution of the virtual workload at the destination host.

20. The method of claim 19, including migrating the virtual workload to the destination host and powering down the source host, based on the carbon footprint associated with the execution of the virtual workload at the destination host being less than the carbon footprint associated with the execution of the virtual workload at the source host.