Allocating cost of disk usage to a linked clone virtual machine based on a parameter of usage

• US 9,582,309 B2
• Filed: 12/09/2014
• Issued: 02/28/2017
• Est. Priority Date: 12/09/2014
• Status: Active Grant

• Current Assignee: Vmware LLC (Broadcom, Inc.)
• Original Assignee: VMware, Inc. (Broadcom, Inc.)
• Inventors: Pannem, Hemanth Kumar, Mirajkar, Akshay Prabhakar
• Primary Examiner(s): Truong, Camquy
• Application Number:
  US14/565,145
• Publication Number:
  US 20160162315A1
• Time in Patent Office: 812 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 2009/45562 Creating, deleting, cloning virtual machine instances
• CPC Class Codes: G06F 2009/45579 I/O management, e.g. providing access to device drivers or storage
• CPC Class Codes: G06F 2009/45583 Memory management, e.g. access or allocation
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

The present disclosure is related to methods, systems, and machine-readable media for allocating cost of disk usage to a linked clone virtual machine (VM) based on a parameter of usage. A determination can be made as to a number of disks used by a linked clone VM among a plurality of disks in a software defined data center over a time period and as to a respective portion of a parameter of usage for each of the number of disks used by the linked clone VM over the time period that is attributable to the linked clone VM. A portion of a cost for usage of each of the number of disks over the time period can be allocated to the linked clone VM in proportion to the respective portion of the parameter of usage attributable to the linked clone VM and/or on a relative latency.

Claims

1. A system, comprising:

processing and memory resources or hardware logic configured to implement;

a disk usage engine configured to;

determine, for a linked clone virtual machine (VM) that is a copy of a parent VM and shares at least one disk with the parent VM, a number of disks used by the linked clone VM among a plurality of disks in a software defined data center over a time period;

determine a respective portion of a parameter of usage for each of the number of disks used by the linked clone VM over the time period that is attributable to the linked clone VM;

build a linked clone tree for a number of VMs that use the plurality of disks during the time period, wherein each node in the tree represents one of the plurality of disks, and

for a leaf in the tree corresponding to the linked clone VM, determine a path from a root of the tree, wherein the path includes a number of nodes representing the number of disks used by the linked clone VM; and

a cost engine configured to allocate, to the linked clone VM, a portion of a cost for usage of each of the number of disks over the time period in proportion to the respective portion of the parameter of usage attributable to the linked clone VM.

2. The system of claim 1, wherein the disk usage engine is configured to poll a hypervisor of the linked clone VM to determine the portion of the parameter of usage, comprising a plurality of input/output operations (IOPs), that is attributable to the linked clone VM.

3. The system of claim 2, wherein the disk usage engine is configured to determine, for each of the plurality of disks, a respective total plurality of IOPs over the time period.

4. The system of claim 3, wherein the plurality of IOPs comprise read IOPs for the plurality of disks.

5. The system of claim 1, wherein the cost engine is configured to allocate the portion of the cost based on a unit rate per unit time.

6. The system of claim 5, wherein the unit rate per unit time is generic to the number of disks.

7. The system of claim 5, wherein the unit rate per unit time is specific for each of the number of disks.

8. The system of claim 1, wherein the cost engine is configured to allocate, to the linked clone VM, the portion of the cost regardless of relative latency of the linked clone VM to latency of other linked clone VMs that use the number of disks.

9. A non-transitory machine-readable medium storing instructions executable by a processing resource to cause a computing system to:

determine, for each of a number of disks used by a linked clone virtual machine (VM) that is a copy of a parent VM and shares at least one disk with the parent VM, among a plurality of disks in a software defined data center, a relative usage attributable to the linked clone VM over a time period in consideration of a relative latency for the linked clone VM over the time period versus other linked clone VMs that use the number of disks;

wherein the instructions to determine the relative usage and the relative latency comprise instructions to determine;

a first value comprising a number of input/output operations (IOPs) attributable to the linked clone VM over a time period in consideration of a latency factor for the linked clone VM over the time period; and

a second value comprising a sum of a number of IOPs attributable to each other linked clone VM that uses the number of disks over the time period in consideration of a latency factor for each other linked clone VM; and

allocate, to the linked clone VM, a cost for usage of each of the number of disks over the time period based on the relative usage and relative latency.

10. The medium of claim 9, wherein the instructions to allocate the cost comprise instructions to allocate, to the linked clone VM, the cost for usage of each of the number of disks over the time period in proportion to a ratio of the first value to the second value.

11. The medium of claim 9, wherein the instructions to determine the relative usage in consideration of the relative latency comprise instructions to determine a latency factor for the linked clone VM.

12. The medium of claim 11, including instructions to determine the latency factor based on:

a latency for the linked clone VM;

an average latency for the linked clone VM and the other linked clone VMs; and

a weight of latency on cost.

13. The medium of claim 12, including instructions to:

measure the latency for the linked clone VM during runtime;

measure the average latency for the linked clone VM and the other linked clone VMs during runtime; and

receive the weight of latency on cost as an input.

14. The medium of claim 9, including instructions to determine the number of disks used by the linked clone VM over the time period.

15. The medium of claim 14, wherein the instructions to determine the number of disks used by the linked clone VM comprise instructions to determine a path from a root of a linked clone tree for the linked clone VM,

wherein the linked clone tree represents a plurality of linked clone VMs that use the plurality of disks during the time period,

wherein each node in the tree represents one of the plurality of disks, and

wherein the path includes a number of nodes representing the number of disks used by the linked clone VM.

16. A method, comprising:

determining a number of disks, among a plurality of disks in a software defined data center, from which snapshots were taken to form a linked clone virtual machine (VM) that is a copy of a parent VM and shares at least one disk with the parent VM;

determining a respective portion of a plurality of input/output operations (IOPs) for the number of disks over a time period that are attributable to the linked clone VM in consideration of a relative latency for the linked clone VM over the time period versus other linked clone VMs that use the number of disks; and

allocating, to the linked clone VM, a portion of a cost for usage of each of the number of disks over the time period in proportion to the respective portion of the plurality of IOPs attributable to the linked clone VM in consideration of the relative latency for the linked clone VM.

17. The method of claim 16, wherein determining the number of disks includes adding a delta disk for the linked clone VM to the number of disks.

18. A non-transitory machine-readable medium storing instructions executable by a processing resource to cause a computing system to:

determine, for each of a number of disks used by a linked clone virtual machine (VM) that is a copy of a parent VM and shares at least one disk with the parent VM, among a plurality of disks in a software defined data center, a relative usage attributable to the linked clone VM over a time period in consideration of a relative latency for the linked clone VM over the time period versus other linked clone VMs that use the number of disks;

wherein the instructions to determine the relative usage in consideration of the relative latency comprise instructions to determine a latency factor for the linked clone VM based on;

a latency for the linked clone VM;

an average latency for the linked clone VM and the other linked clone VMs; and

a weight of latency on cost;

measure the latency for the linked clone VM during runtime;

measure the average latency for the linked clone VM and the other linked clone VMs during runtime;

receive the weight of latency on cost as an input; and

allocate, to the linked clone VM, a cost for usage of each of the number of disks over the time period based on the relative usage and relative latency.

19. A non-transitory machine-readable medium storing instructions executable by a processing resource to cause a computing system to:

determine, for each of a number of disks used by a linked clone virtual machine (VM) that is a copy of a parent VM and shares at least one disk with the parent VM, among a plurality of disks in a software defined data center, a relative usage attributable to the linked clone VM over a time period in consideration of a relative latency for the linked clone VM over the time period versus other linked clone VMs that use the number of disks;

determine the number of disks used by the linked clone VM over the time period by determining a path from a root of a linked clone tree for the linked clone VM,

wherein the linked clone tree represents a plurality of linked clone VMs that use the plurality of disks during the time period,

wherein each node in the tree represents one of the plurality of disks, and

wherein the path includes a number of nodes representing the number of disks used by the linked clone VM; and

allocate, to the linked clone VM, a cost for usage of each of the number of disks over the time period based on the relative usage and relative latency.