Browse Prior Art Database

Technique for dynamically allocating resources in a heterogenous hypervisor environment for a given hardware architecture.

IP.com Disclosure Number: IPCOM000236199D
Publication Date: 2014-Apr-11

Publishing Venue

The IP.com Prior Art Database

Abstract

Technique for dynamically allocating resources in a heterogenous hypervisor environment for a given hardware architecture.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 49% of the total text.

Page 01 of 12

Technique for dynamically allocating resources in a heterogenous hypervisor environment for a given hardware architecture .


1. It is common these days in a typical enterprise environment to Virtualize the hardware resources so as to make the most efficient use of hardware in a data center.


2. The hardware architecture could be uniform such as Intel / AMD x86 or have a heterogenous architecture mix, such as IBM Power, Sun Solaris etc. From here on we are going to consider all scenarios for a given hardware architecture only.


3. It is also increasingly the case that the Hypervisors are not the same even if the hardware architecture is the same. Heterogenous Hypervisors are needed to support specific operating systems and specific workloads.

For eg, Hyper-V could be used for virtualizing a solution that pre dominantly involves Windows servers,

KVM could be used to virtualize a solution that pre dominantly involves Linux Servers,

and VMware could be used where both Linux and Windows servers are part of a solution.


4. These virtualized solutions usually consist of a cluster of servers that are homogenous from a Hypervisor perspective.


5. Clusters are architected based on current capacity and expected capacity growth and accordingly given hardware resources.


6. This could lead to some clusters being over provisioned with capacity and some others with lesser capacity based on growth projections.


7. This is not an ideal way to use the physical resources in a dynamic environment where ideally capacity needs to be provisioned to the right cluster on demand and de-provisioned when no longer needed.


8. Currently there is no way to achieve this in an automated fashion when multiple Hypervisors are involved and resources need to be dynamically moved to the cluster as needed.


9. The proposed solution details a technique of provisioning a given node to the most appropriate cluster regardless of the Hypervisor that is used to manage that cluster.

This disclosure talks about how to deploy hardware resources in a dynamic and automated way in a heterogenous Hypervisor environment.

Customer data centers are increasingly having to deal with multiple Hypervisors. This disclosure automates the management of hardware resources in a heterogenous Hypervisor environment. (Hardware, Server and Node are used inter changeable for this discussion)

Advantages:
1. Customers no longer need to worry about manually installing and load balancing hardware across Hypervisor clusters.

2. During the course of operation of a cluster, there are times when the cluster can be highly loaded with respect to the hardware resources and at other times, lightly loaded. By dynamically moving the hardware resources where there is the most need, this disclosure, reduces the manual overheads involved.

3. It also makes the most efficient use of resources across all clusters in operation.

4. This results in overall power and cooling costs as low resource usage clusters are compacted and...