Optimized resource allocation and management in a virtualized computing environment

Patent metadata

Field	Value
Publication number	US-9405572-B2
Application number	US-201414246148-A
Country	US
Kind code	B2
Filing date	Apr 7, 2014
Priority date	Apr 7, 2014
Publication date	Aug 2, 2016
Grant date	Aug 2, 2016

Abstract

Machines, systems and methods for allocating resources to in a virtualized computing environment, the method comprising detecting one or more host machines with resources allocated to one or more virtual machines (VMs) that are in an idle state; reducing resource entitlements for at least one of the VMs that is detected to be in the idle state to make more resources available for allocation to VMs that are not in the idle state; and increasing resource entitlements for at least one of the VMs with reduced entitlement, in response to determining that the VM with reduced entitlement is no longer in the idle state.

First claim

What is claimed is:

1. A method for allocating resources in a virtualized computing environment, the method comprising:
detecting, by an idleness detector, one or more host machines with resources allocated to one or more virtual machines (VMs) that are in an idle state;
reducing, by the idleness detector triggering a throttling event, resource entitlements for at least one of the VMs that is detected to be in the idle state to make more resources available for allocation to VMs that are not in the idle state;
increasing resource entitlements for at least one of the VMs with reduced entitlement, in response to determining that the VM with reduced entitlement is no longer in the idle state,
wherein a throttling scheduler manages the throttling event to enable admission of additional VMs on an identified host machine for the purpose of resource allocation,
wherein the admission of the additional VMs on the identified host machine is based on a probability of the additional VMs not causing the identified host machine to run out of allocated resources according to the following algorithm:
D i =μ i +σ i ·Z p i

D i is an equivalent demand for VMs of type i, μ i is an average demand for VMs of type i, σ i is a standard deviation of demand for VMs of type i, and Z is a p i quantile of unit normal distribution N(0,1); and
recording in a throttling history the placement of the additional VMs on the identified host machine.

2. The method of claim 1 , wherein an admission controller manages the admission of new VMs into the virtualized computing environment, in response to determining that at least one or more hosts have free resources to allocate to a new VM.

3. The method of claim 2 , wherein the idleness detector identifies that a VM is no longer in the idle state based on determining that an idle VM is highly utilizing remaining resources allocated to the VM.

4. The method of claim 3 , wherein resource starvation is manifested by high resource utilization.

5. The method of claim 4 , wherein the throttling scheduler attempts to restore an earlier entitlement of resources for an idle VM with reduced entitlement, in response to determining that the VM is starving for resources.

6. The method of claim 5 , wherein the starving VM is migrated from a first host to a second host, in response to determining that the earlier entitlement for resources is not restorable.

7. The method of claim 6 , wherein the admission controller prevents allocation of additional resources to any VM, in response to determining that no host is present with available resources to accommodate the migration of the starving VM.

8. A system for allocating resources in a virtualized computing environment, the system comprising:
a processor; and
a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by the processor to cause the system to:
detect, by an idleness detector, one or more host machines with resources allocated to one or more virtual machines (VMs) that are in an idle state;
reduce, by the idleness detector triggering a throttling event, resource entitlements for at least one of the VMs that is detected to be in the idle state to make more resources available for allocation to VMs that are not in the idle state;
increase resource entitlements for at least one of the VMs with reduced entitlement, in response to determining that the VM with reduced entitlement is no longer in the idle state,
wherein a throttling scheduler manages the throttling event to enable admission of additional VMs on an identified host machine for the purpose of resource allocation,
wherein the admission of the additional VMs on the identified host machine is based on a probability of the additional VMs not causing the identified host machine to run out of allocated resources according to the following algorithm:
D i =μ i +σ i ·Z p i

D i is an equivalent demand for VMs of type i, μ i is an average demand for VMs of type i, σ i is a standard deviation of demand for VMs of type i, and Z is a p i quantile of unit normal distribution N(0,1); and
record in a throttling history the placement of the additional VMs on the identified host machine.

9. The system of claim 8 , wherein an admission controller manages the admission of new VMs into the virtualized computing environment, in response to determining that at least one or more hosts have free resources to allocate to a new VM.

10. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, wherein the computer readable storage medium is not a transitory signal per se, the program instructions executable by a computer to perform a method for allocating resources in a virtualized computing environment, the method comprising:
detecting, by an idleness detector, one or more host machines with resources allocated to one or more virtual machines (VMs) that are in an idle state;
reducing, by the idleness detector triggering a throttling event, resource entitlements for at least one of the VMs that is detected to be in the idle state to make more resources available for allocation to VMs that are not in the idle state;
increasing resource entitlements for at least one of the VMs with reduced entitlement, in response to determining that the VM with reduced entitlement is no longer in the idle state,
wherein a throttling scheduler manages the throttling event to enable admission of additional VMs on an identified host machine for the purpose of resource allocation,
wherein the admission of the additional VMs on the identified host machine is based on a probability of the additional VMs not causing the identified host machine to run out of allocated resources according to the following algorithm:
D i =μ i +σ i ·Z p i

D i is an equivalent demand for VMs of type i, μ i is an average demand for VMs of type i, σ i is a standard deviation of demand for VMs of type i, and Z is a p i quantile of unit normal distribution N(0,1); and
recording in a throttling history the placement of the additional VMs on the identified host machine.

11. The computer program product of claim 10 , wherein an admission controller manages the admission of new VMs into the virtualized computing environment, in response to determining that at least one or more hosts have free resources to allocate to a new VM.

Assignees

• IBM

Inventors

• Breitgand David
• Dubitzky Zvi
• Epstein Amir
• Feder Oshrit
• Glikson Alexander
• Kravchenco Pavel
• Shapira Inbar
• Toffetti-Carughi Giovanni

Classifications

• G06F9/45558 Primary

  Hypervisor-specific management and integration aspects · CPC title

• G06F2009/45591

  Monitoring or debugging support · CPC title

• G06F2009/45595

  Network integration; Enabling network access in virtual machine instances · CPC title

• G06F2009/4557

  Distribution of virtual machine instances; Migration and load balancing · CPC title