Browse Prior Art Database

Placement of Constrained Groups of Logical Application Components in a Cloud Computing Environment

IP.com Disclosure Number: IPCOM000240632D
Publication Date: 2015-Feb-13
Document File: 8 page(s) / 101K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method to provide an end-to-end solution for a Cloud computing environment that produces an understanding of application and infrastructure requirements, places an application to meet those requirements, and then optimizes a response to changes in the system. The method uses constrained groups of logical application components to resolve the placement of multiple tiers in an application.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 38% of the total text.

Page 01 of 8

Placement of Constrained Groups of Logical Application Components in a Cloud Computing Environment

In Cloud environments, most applications have deployment patterns with multiple tiers, between which are unique placement requirements. For example, an instance of tier A depends on service provided by an instance of tier B running in the same server ; instances of tier B require even distribution among multiple availability zones for High Availability (HA) purposes, while keeping network distance as close as possible for better performance. Meanwhile, the infrastructure level must fully utilize physical resources to minimize the energy and maintenance costs.

Existing methods and algorithms deal with placing individual virtual machines on physical machines, and only focus on individual instances. These approaches cannot resolve the placement of multiple tiers in an application .

When a collection of virtual machines were considered as an application to be placed in the cloud , along with communication requirements among them, the problem became more complex, since the physical communication paths had to satisfy any virtual bandwidth requirements. Past methods or systems to address this only focus on part of the placement requirements of an application, such as the communication bandwidth between instances. Those approaches do not consider all of the requirements of the application along with the infrastructure level requirements.

A method is needed to provide an end-to-end solution that produces an understanding of application and infrastructure requirements, places an application to meet those requirements, and then optimizes a response to changes in the system.

The novel approach is to describe the placement requirements of an application by :

Having an intra -tier policy Having an inter -tier policy Describing infrastructure level requirements by infrastructure policy
Having a schedule engine to place all instances of an application holistically to meet both the application policy and the

  infrastructure policy
Providing an optimization service to optimize server distribution as the response to the system change by querying the

schedule engine for an optimization plan

With this approach, the scheduler considers a general, holistic objective function that includes an infrastructure component (e.g., load balancing or consolidation among physical nodes) as well as an application (i.e. workload) component that captures whether

1


Page 02 of 8

the placement of the workload is good, beyond just meeting requirements. In addition to networking requirements, as only considered in previous approaches, the novel method deals with node resource requirements, shared resources such as licenses, and other edge constraints such as availability constraints through affinity and anti-affinity relations.

The novel method has four parts. The first part manages the application and allows the user to specify the resources and placement...