PARALLEL DATA LAB 

PDL Abstract

A Comparative Study of Baremetal Provisioning Frameworks

Carnegie Mellon University Parallel Data Lab Technical Report CMU-PDL-14-109. December 2014.

Ashok Chandrasekar, Garth A. Gibson

Carnegie Mellon University

http://www.pdl.cmu.edu/

There are many baremetal frameworks available in the market today. These baremetal frameworks help ease the work of a cluster/datacenter administrator by automating the deployment of operating systems and other software onto the individual machines in a datacenter. These frameworks differ from each other on several aspects like price, stability, maintainability, feature support and performance. As a result, it becomes difficult to choose the best framework to suit one’s needs. This study aims at helping the administrator in choosing the appropriate framework based on their needs by providing a comparison of these frameworks with a main emphasis on Emulab and OpenStack Ironic.

KEYWORDS: baremetal frameworks, network simulators, scheduling, simulated annealing, emulab, ironic

FULL TR: pdf