2015Unpublished venueRequires access

Pack Up Cloud: Recursive Datacenter Resource Management and Experimental Studies

Ye Xia, Shigang Chen, Maurício Tsugawa, J.A.B. Fortes

Open publisher page 0 citations

Abstract

Today's virtualized resource management in datacenters is typically VM-centric, which has serious problems in scalability. We propose a shift from the traditional flat, fine-grained model towards a hierarchical, abstract model that facilitates decomposition of the complex resource allocation problem in a divide-and-conquer approach. In particular, this paper elaborates a pack-centric framework in datacenter resource management, provides a case study with a recursive algorithm, and presents detailed results from experimental studies on VM-PM mapping through recursive resource allocation, which demonstrates the performance and scalability of the pack-centric framework.

About this research paper

What this paper is about

Today's virtualized resource management in datacenters is typically VM-centric, which has serious problems in scalability. We propose a shift from the traditional flat, fine-grained model towards a hierarchical, abstract model that facilitates decomposition of the complex resource allocation problem in a divide-and-conquer approach. In particular, this paper elaborates a pack-centric framework in datacenter resource management, provides a case study with a recursive algorithm, and presents detailed results from experimental studies on VM-PM mapping through recursive resource allocation, which demonstrates the performance and scalability of the pack-centric framework.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Today's virtualized resource management in datacenters is typically VM-centric, which has serious problems in scalability. We propose a shift from the traditional flat, fine-grained model towards a hierarchical, abstract model that facilitates decomposition of the complex resource allocation problem in a divide-and-conquer approach. In particular, this paper elaborates a pack-centric framework in datacenter resource management, provides a case study with a recursive algorithm, and presents detailed results from experimental studies on VM-PM mapping through recursive resource allocation, which demonstrates the performance and scalability of the pack-centric framework.

Key concepts: Scalability, Computer science, Cloud computing, Distributed computing, Resource management (computing), Resource allocation, Resource (disambiguation), Decomposition

Related papers

Back to paper searchBrowse research topicsOriginal source
Pack Up Cloud: Recursive Datacenter Resource Management and Experimental Studies — Research Paper | ScholarLens