Policy-driven capacity management in resource provisioning environments

• US 8,732,310 B2
• Filed: 04/22/2010
• Issued: 05/20/2014
• Est. Priority Date: 04/22/2010
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Breitgand, David, Rochwerger, Benny, Satran, Julian
• Primary Examiner(s): Tiv, Backhean
• Application Number:
  US12/764,987
• Publication Number:
  US 20110264805A1
• Time in Patent Office: 1,489 Days
• Field of Search: 709224-226, 718/1, 370/229
• US Class Current: 709/226
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

Systems and methods for policy-driven capacity management in a resource provisioning environment, the method comprising storing, in an operational database one or more virtual resource sets (VRSs) and elasticity ranges for components to be deployed in a resource provisioning environment, and probabilistic guarantees on the elasticity ranges defined in a service manifest provided by a service subscriber; collecting historical data about capacity usage in the resource provisioning environment and failure statistics from the operational database, in addition to prior service commitments due to previously contracted service level agreements (SLAs) stored in a SLA repository; calculating equivalent capacity for the resource provisioning environment based on a defined residual benefit goal, and other business goals instrumented by way of a policy engine; and placing requested virtual resources on physical resources, in response to determining that sufficient physical capacity is available to host the calculated equivalent capacity.

Claims

1. A computer-implemented system for policy-driven capacity management in a resource provisioning environment, the system comprising:

an admission controller for storing, in an operational database;

one or more virtual resource sets (VRSs) and elasticity ranges for components to be deployed in the resource provisioning environment, and

probabilistic guarantees on the elasticity ranges defined in a service manifest provided by a service subscriber;

a monitor for collecting historical data about capacity usage in the resource provisioning environment and failure statistics from the operational database, in addition to prior service commitments due to previously contracted service level agreements (SLAs) stored in a SLA repository;

an equivalent capacity calculator for calculating equivalent capacity for the resource provisioning environment based on a defined residual benefit goal, and other business goals instrumented by way of a policy engine,

wherein equivalent capacity refers to minimum allocated capacity sufficient to guarantee QoS commitments subject to acceptable risk level of non-compliance due to congestion, such that equivalent capacity is as close to physical capacity as possible; and

a placement controller for placing requested virtual resources on physical resources, in response to determining that sufficient physical capacity is available to host the calculated equivalent capacity,

wherein an acceptable risk level (ARL) with respect to SLA non-compliance due to congestion for a service provider is defined by a policy engine such that if the available physical capacity is greater than the equivalent capacity and feasible placement of virtual resources belonging to the VRSs exists for an equivalent capacity for resource ranges defined for a usage windows, then availability SLAs of VRSs are protected at with probability=1−

ARL, where ARL is the acceptable risk level probability.

2. The system of claim 1, wherein the admission controller enforces a rejection handling policy, in response to determining that the available physical capacity is insufficient to host the calculated equivalent capacity.

3. The system of claim 1, wherein the admission controller admits a requested service by the service subscriber, in response to determining that sufficient physical capacity is available to host the calculated equivalent capacity.

4. The system of claim 1 further comprising a capacity planning calculator periodically invoked to calculate expected equivalent capacity for different time frames based on data that suggests use trends and other relevant information collected and stored in the operational database and the SLA repository.

5. The system of claim 4, wherein the capacity planning calculator notifies a capacity planner that additional physical capacity is needed to cover the calculated equivalent capacity, in response to determining that the physical capacity is insufficient to host the calculated equivalent capacity.

6. The system of claim 1 wherein equivalent capacity approximation is utilized in context of the bandwidth capacity planning where the set (C1, C2, . . . , Cn)tv represents a vector of random variables such that physical machines of each type deployed by a VRS at time instance “

t”

where the maximal number of VMs of each type that may be deployed by each VRS is bound by a range agreed upon in the corresponding service level objectives (SLO) clause of an SLA for the VRS.

7. The system of claim 6 wherein the VRSs behave independently and are reasonably well-approximated using Gaussian approximation according to the formula:

Ci(ARL)=Mi+Si*(SQRT(−

21n(ARL)−

1n(2PI)), where Mi=SUMv Mvi (total mean in standardized capacity units) and Si=SUMv Svi (total standard deviation).

8. The system of claim 7 wherein if a VM of each type constitutes a standard capacity unit, then capacity requirements of a VRS as a whole may be represented as a vector of pairs CR=({C1 min,C1max}, {C2 min,C2max}, . . . , {Cnmin,Cnmax}), where Cjmin stands for the minimal number of standard capacity units of type j and Cjmax stands for the maximal units thereof and a pair {Cjmin,Cjmax} comprises a capacity range for VMs of type j expressed in the standard capacity units for this type.

9. The system of claim 8 wherein an availability SLOs for a VRS may be generalized as follows:

with probability P, capacity range CR=({C1 min,C1max}, {C2 min,C2max}, . . . , {Cnmin,Cnmax}) of the VRS will be available throughout usage window W where probability P is computed over SLA evaluation period E.

10. A computer program product comprising a non-transitory computer readable storage medium having a computer readable program, wherein the computer readable program when executed on a computer causes the computer to:

store, in an operational database;

one or more virtual resource sets (VRSs) and elasticity ranges for components to be deployed in a resource provisioning environment, and

probabilistic guarantees on the elasticity ranges defined in a service manifest provided by a service subscriber;

collect historical data about capacity usage in the resource provisioning environment and failure statistics from the operational database, in addition to prior service commitments due to previously contracted service level agreements (SLAs) stored in a SLA repository;

calculate equivalent capacity for the resource provisioning environment based on a defined residual benefit goal, and other business goals instrumented by way of a policy engine, wherein equivalent capacity refers to minimum allocated capacity sufficient to guarantee QoS commitments subject to acceptable risk level of non-compliance due to congestion, such that equivalent capacity is as close to physical capacity as possible; and

place requested virtual resources on physical resources, in response to determining that sufficient physical capacity is available to host the calculated equivalent capacity,

wherein an acceptable risk level (ARL) with respect to SLA non-compliance due to congestion for a service provider is defined by a policy engine such that if the available physical capacity is greater than the equivalent capacity and feasible placement of virtual resources belonging to the VRSs exists for an equivalent capacity for resource ranges defined for a usage windows, then availability SLAs of VRSs are protected at with probability=1−

ARL, where ARL is the acceptable risk level probability.

11. The computer program product of claim 10, wherein the computer readable program when executed on a computer further causes the computer to enforce a rejection handling policy, in response to determining that the available physical capacity is insufficient to host the calculated equivalent capacity.

12. The computer program product of claim 10, wherein the computer readable program when executed on a computer further causes the computer to admit a requested service by the service subscriber, in response to determining that sufficient physical capacity is available to host the calculated equivalent capacity.

13. The computer program product of claim 10, wherein an equivalent capacity calculator is periodically invoked to calculate expected equivalent capacity for different time frames based on data that suggests use trends and other relevant information collected and stored in the operational database and the SLA repository.

14. A computer-implemented method for policy-driven capacity management in a resource provisioning environment, the method comprising:

storing in an operational database;

one or more virtual resource sets (VRSs) and elasticity ranges for components to be deployed in the resource provisioning environment, and

probabilistic guarantees on the elasticity ranges defined in a service manifest provided by a service subscriber;

collecting historical data about capacity usage in the resource provisioning environment and failure statistics from the operational database, in addition to prior service commitments due to previously contracted service level agreements (SLAs) stored in a SLA repository;

calculating equivalent capacity for the resource provisioning environment based on a defined residual benefit goal, and other business goals instrumented by way of a policy engine,

wherein equivalent capacity refers to minimum allocated capacity sufficient to guarantee QoS commitments subject to acceptable risk level of non-compliance due to congestion, such that equivalent capacity is as close to physical capacity as possible; and

placing requested virtual resources on physical resources, in response to determining that sufficient physical capacity is available to host the calculated equivalent capacity,

wherein an acceptable risk level (ARL) with respect to SLA non-compliance due to congestion for a service provider is defined by a policy engine such that if the available physical capacity is greater than the equivalent capacity and feasible placement of virtual resources belonging to the VRSs exists for an equivalent capacity for resource ranges defined for a usage windows, then availability SLAs of VRSs are protected at with probability=1−

ARL, where ARL is the acceptable risk level probability.

15. The method of claim 14, wherein a rejection handling policy is enforced, in response to determining that the available physical capacity is insufficient to host the calculated equivalent capacity.

16. The method of claim 14, wherein a requested service by the service subscriber is admitted, in response to determining that sufficient physical capacity is available to host the calculated equivalent capacity.

17. The method of claim 14 further comprising periodically calculating expected equivalent capacity for different time frames based on data that suggests use trends and other relevant information collected and stored in the operational database and the SLA repository.

18. The method of claim 14, further comprising notifying that additional physical capacity is needed to cover the calculated equivalent capacity, in response to determining that the physical capacity is insufficient to host the calculated equivalent capacity.