Browse Prior Art Database

A Method and Apparatus for Online Application Placement in Hierarchical Cloud Computing Environments

IP.com Disclosure Number: IPCOM000241320D
Publication Date: 2015-Apr-16
Document File: 4 page(s) / 109K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is an application placement method that is both practical and online, and provides performance guarantees for placement of hierarchical application graphs in hierarchical cloud computing environments.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 36% of the total text.

Page 01 of 4

A Method and Apparatus for Online Application Placement in Hierarchical Cloud Computing Environments

This article is related to the field of cloud computing , specifically focusing on a method for online application placement in hierarchical cloud computing environments .

Application placement mechanisms in cloud computing systems map application

workloads to cloud physical resources subject to resource constraints in order to optimize cloud system performance (e.g., minimize resource consumption, reduce delay, balance load, etc.). Application placement is a crucial component of a cloud computing system because efficient allocation of resources yields efficient operation and enhanced user experience. Furthermore, it has direct implications in the cost of operating the data center and the charging costs to the users .

Application placement has been extensively addressed in the context of deployment of

Virtual Machines (VM) in data center clusters, which are a representative instantiation of a modern cloud computing system. Most techniques for mapping VMs to data centers use bin packing heuristics, where each VM is mapped to physical resources according to the associated Central Processing Unit (CPU), communication Input/Output (I/O), and memory requirements. VM workloads have multi-dimensional CPU, I/O, and memory requirements and place those to processors subject to the associated CPU, I/O, and memory resources. Such methods only take into account processing resources and view VMs as single computation entities. Those methods also implicitly or explicitly assume that the processors of the data center clusters are homogeneous, and do not take into account the communication interaction between

VMs or the link bandwidth between the processors.

Cloud computing systems are becoming increasingly more complex. For example, data centers are becoming more heterogeneous in communication and computing resources and the processors can be distributed in multiple geographical locations . Furthermore, the application workloads running in cloud computing systems become increasingly complex (modeling processing interactions) and more diverse (i.e. not just VMs) over time.

Therefore, practicalmethods that cope with heterogeneous cloud computing environments and complex workloads become imperative. There is a need for application placement techniques that are practical , online, and that go beyond heuristics by providing performance guarantees with respect to the optimal solution .

The novel contribution is an application placement method with provable performance guarantees from the optimal that can be deployed in cloud computing systems that arise in practice. More specifically, the solution provides a practical and online method that provides performance guarantees for placement of hierarchical application graphs in hierarchical cloud computing environments.

This approach includes an exact optimal placement technique for linear application grap...