Integration of PXE-Based Desktop Solutions Into Broadband Access Networks

Total Page:16

File Type:pdf, Size:1020Kb

Integration of PXE-Based Desktop Solutions Into Broadband Access Networks Integration of PXE-based Desktop Solutions into Broadband Access Networks Tiago Cruz1, Paulo Simões1, Fernando Bastos2, Edmundo Monteiro1 1DEI-CISUC, University of Coimbra 2PT Inovação Coimbra, Portugal Aveiro, Portugal {tjcruz, psimoes, edmundo}@dei.uc.pt [email protected] Abstract — Presently there is a lack of remote desktop user data and access applications in the cloud. Netbooks, management solutions for domestic and SOHO users connected tablets and some smartphone platforms are now specifically to broadband access networks. This contrasts with the enterprise designed as thin computing devices. LAN environment, where there are several standards, resources At the present evolution stage, even if many users could and frameworks for PC or thin-client management. Among these, already permanently live and work on a cloud environment, one specific remote boot technology – the Preboot eXecution Environment (PXE) [1] – is now the basis for a wide array of traditional applications are still the norm on many usage LAN-wide desktop management applications. categories. As an example, even if some office and In this context, integrating PXE-based solutions into productivity suites are already offered as a service (e.g. broadband access networks would allow novel management Google Docs [2], Zoho [3]), they remain somewhat limited in paradigms, targeting not just domestic end-users but also comparison to their traditional, locally deployed counterparts, telecommuters working from their homes and small businesses in terms of features, functionality and usability. Also, some which are too small for local deployment of full-fledged traditional applications are starting to embed support for cloud enterprise desktop management platforms. service components (e.g. Microsoft Office suite [4]) but In this paper we propose a solution that brings the benefits of without replacing the traditional desktop computing model by managed desktop computing to home users, telecommuters and cloud computing. Instead, these two approaches will likely small businesses by integrating PXE technologies into broadband coexist, cooperate and merge with each other. access network environments. In addition, we also propose a The mainstream desktop device is still the standalone PC, desktop services delivery model capable of efficiently providing a secure and quality managed desktop experience to domestic and with a Total Cost of Ownership (TCO) that may largely SOHO end-users, using a PXE-based thin-client platform for exceed its initial acquisition cost, when including maintenance broadband environments that can replace a full-fledged PC and indirect costs. Organizations with dozens, hundreds or whilst maintaining most of its benefits. thousands of PCs feel this problem in a much bigger scale. For them, the industry created specific standards and tools for Keywords — PXE, Access Networks, CWMP/TR-069, DaaS enhanced desktop management. Managed PCs and thin-clients are both a result of those efforts. I. INTRODUCTION Yet, most of those technologies were developed for Regarded as the logical evolution beyond the centralized corporate LAN environments, leaving out domestic, Small (mainframe-based) and client-server paradigms, cloud Office Home Office (SOHO) users or even small computing is a somewhat vague term (more of a metaphor) organizations served by commodity broadband Internet that encompasses a wide array of technologies and concepts services. For these situations, existing alternatives are limited: that work together to allow the delivery and consumption of • Unmanaged standalone PCs, with a significant TCO services hosted and supported by remote data centers overhead and unable to be remotely diagnosed or recovered (providing dynamically scalable and often virtualized from bare metal, in case of critical failures. computing resources) to another service or an end-user, • generally using a web browser as a universal client. Standalone Intel vPro–certified [5] PCs with embedded out- of-band desktop management capabilities. However, vPro However, the fundamental cloud computing concept of is targeted towards conventional PCs and is only available delivering everything-as-a-service is heavily dependent on the on limited hardware and firmware combinations. existence of reliable and capable data pipes connecting • providers to service consumers. As such, cloud computing In the case of telecommuters or remote branches of larger owes much of its success to the increasingly available high- organizations, the use of remote desktop protocols – such as speed commodity broadband access networks (fixed and Microsoft’s Remote Desktop Protocol (RDP) [6], Citrix mobile) without which it would be an unfeasible proposition. Independent Computing Architecture (ICA) [7] or Virtual Network Computing (VNC) [8] – over SSH [9], SSL/TLS Also, the widespread availability of broadband network [10] or VPN tunnels on PCs or thin-clients. However, access, together with cloud services, spawned a new breed of depending on the VPN technology, thin clients can be thin-computing devices that heavily rely on such services. tricky to configure – alternatively the VPN client could be Instead of using the traditional model – where data and configured on the broadband router, an operation that might applications reside on the device itself – those devices store also require technical skills lacked by most users. • Web appliances. However, despite recent developments, these appliances are still unsuitable for many applications and may represent, by themselves, a management problem. Part of these problems might be addressed using remote boot Figure 1: PXE boot ROM API. technology. Until recently, network boot over access networks PXE-compliant Boot ROMs provide the means to control the was unfeasible due to bandwidth limitations. However, with boot process in order to download and execute either a full- broadband access networks bandwidth steadily increasing, this blown OS or just a small pre-boot management agent for restriction is disappearing, shifting the focus to the remaining diagnostic or pre-staging purposes. Through the use of PXE it obstacles – like the fact that PXE [1], the standard remote is possible to configure a desktop PC boot sequence to be network boot protocol, uses mechanisms such as the Dynamic preceded by a PXE boot attempt before using local mass Host Configuration Protocol (DHCP [11]) and the Trivial File storage devices, making it possible to download and boot a Transfer Protocol (TFTP [12]) in ways that make it unsuitable remote OS or a remote agent (in order to initiate maintenance for naked use over access networks or WAN links. tasks) or to proceed with the normal boot sequence from local In this paper we propose a solution for integrating PXE storage. Alternatively, it is possible to configure PXE boot to technology on broadband access networks, in order to allow be attempted only in case of local storage device failure (as a for better management of existing desktops and to enable the recovery mechanism). The operation of a PXE boot ROM creation of completely stateless thin-client devices capable of follows a simple three-stage process (Figure 2): securely booting a remote OS over broadband links. • IP subsystem initialization. The PXE Boot ROM gets a First we present a solution to overcome PXE limitations on valid IP via the DHCP protocol (1,2), together with DHCP access networks. In this perspective we also discuss how to option tags that identify the presence of PXE support at the integrate PXE support in the Internet Service Provider (ISP) DHCP server level, together with the location of the TFTP management infrastructure – making use of the CPE WAN server and the file name of the NBP to be downloaded. Management Protocol (CWMP) [13] to control PXE-related • TFTP download of the NBP (3,4), using TFTP. service parameters on the user’s access router. Next we • discuss application models for this solution, presenting Execution of the downloaded NBP agent (5,6,7). specific use cases where it can be used to remotely manage desktops, providing boot support for install, recovery or update procedures. We also discuss how to implement a completely stateless thin-client – thus enabling a complete end-to-end Desktop-as-a-Service model (DaaS) based on remote desktop technologies and boot-time downloaded OS. The rest of this paper is organized as follows. Section 2 discusses the PXE protocol and its usage on access networks. Section 3 discusses how to integrate CWMP-based access network management with PXE-enabled desktop management. Section 4 addresses application scenarios and Section 5 discusses implementation and validation. Section 6 Figure 2: PXE Agent Download Process. discusses related work and Section 7 concludes the paper. B. The problem with PXE on access networks When originally conceived, using PXE outside LAN II. INTEGRATION OF PXE ON ACCESS ENVIRONMENTS environments was not envisaged, since the download latency PXE is a Network Boot firmware extension for PC BIOS would be too high, even for small boot agents with limited created in the context of the Intel Boot Initiative. Supported by functionality. Meanwhile this assumption has been challenged most Network Interface Cards, it is a de facto standard for by broadband Internet access technologies like Asymmetric network boot. It was originally conceived as a special piece of Digital Subscriber Line (ADSL) and Gigabit Passive Optical firmware (the PXE boot ROM) that allowed to use the network Networks (GPON)
Recommended publications
  • UEFI PXE and Ipxe Alternative Approaches to PXE Booting
    Installing ESXi Using PXE n gPXELINUX is a hybrid configuration that includes both PXELINUX and gPXE and supports booting from a Web server. gPXELINUX is part of the SYSLINUX package. If you use gPXELINUX to boot the ESXi installer, only the gpxelinux.0 binary file, mboot.c32, and the configuration file are transferred via TFTP. The remaining files are transferred via HTTP. HTTP is typically faster and more reliable than TFTP, especially for transferring large amounts of data on a heavily loaded network. NOTE VMware currently builds the mboot.c32 plugin to work with SYSLINUX version 3.86 and tests PXE booting only with that version. Other versions are likely to be incompatible. This is not a statement of limited support. For support of third-party agents that you use to set up your PXE booting infrastructure, contact the vendor. UEFI PXE and iPXE Most UEFI firmware natively includes PXE support that allows booting from a TFTP server. The firmware can directly load the ESXi boot loader for UEFI systems, mboot.efi. Additional software such as PXELINUX is not required. iPXE can also be useful for UEFI systems that do not include PXE in firmware and for older UEFI systems with bugs in their PXE support. For such cases you can try installing iPXE on a USB flash drive and booting from there. NOTE Apple Macintosh products do not include PXE boot support. They include support for network booting via an Apple-specific protocol instead. Alternative Approaches to PXE Booting Alternative approaches to PXE booting different software on different hosts are also possible, for example: n Configuring the DHCP server to provide different initial boot loader filenames to different hosts depending on MAC address or other criteria.
    [Show full text]
  • Project Report - Adding PXE Boot Into Palacios
    Project Report - Adding PXE Boot into Palacios Chen Jin Bharath Pattabiraman Patrick Foley EECS Department EECS Department EECS Department Northwestern University Northwestern University Northwestern University chen.jin@eecs. bharath@u. patrickfoley2011@u. northwestern.edu northwestern.edu northwestern.edu ABSTRACT PXE is a standard for booting an OS from the network. Most machines BIOSes support it. But, the BIOS used by Palacios guests did not. In our project, we tried various ways in which PXE network boot capability could be added to Palacios. We used a PXE-capable Etherboot ROM image from ROM-o-matic.net that has support for our emulated network card. We then used this small ISO image to build the guest and let it serve as a replacement PXE-boot ROM for the emulated network card. With passthrough I/O, the requests are handed over directly to the host, which are then sent to the DHCP and Boot servers to initiate the network boot process. The PXE capability will of vital importance in diskless nodes where the node is completely dependent on Figure 1: PXE system configuration the network for booting. 1. INTRODUCTION using PXE protocol and then boots the guest. PXE (Preboot eXecution Environment) allows us to boot Kitten/Palacios (and a test guest) remotely from a network server. Booting Palacios/Kitten over a network server is 2. SYSTEM already possible. In this research effort we have enabled So, as shown in Figure 1, in order to use PXE we need to Palacios to remote boot a guest OS using PXE. setup a PXE-server which can allow client systems to: PXE is defined on a foundation of Internet protocols, namely • TCP/IP, DHCP, and TFTP.
    [Show full text]
  • Vmware Esxi Installation and Setup
    VMware ESXi Installation and Setup 02 APR 2020 Modified on 11 AUG 2020 VMware vSphere 7.0 VMware ESXi 7.0 VMware ESXi Installation and Setup You can find the most up-to-date technical documentation on the VMware website at: https://docs.vmware.com/ VMware, Inc. 3401 Hillview Ave. Palo Alto, CA 94304 www.vmware.com © Copyright 2018-2020 VMware, Inc. All rights reserved. Copyright and trademark information. VMware, Inc. 2 Contents 1 About VMware ESXi Installation and Setup 5 Updated Information 6 2 Introduction to vSphere Installation and Setup 7 3 Overview of the vSphere Installation and Setup Process 8 4 About ESXi Evaluation and Licensed Modes 11 5 Installing and Setting Up ESXi 12 ESXi Requirements 12 ESXi Hardware Requirements 12 Supported Remote Management Server Models and Firmware Versions 15 Recommendations for Enhanced ESXi Performance 15 Incoming and Outgoing Firewall Ports for ESXi Hosts 17 Required Free Space for System Logging 19 VMware Host Client System Requirements 20 ESXi Passwords and Account Lockout 20 Preparing for Installing ESXi 22 Download the ESXi Installer 22 Options for Installing ESXi 23 Media Options for Booting the ESXi Installer 24 Using Remote Management Applications 35 Customizing Installations with vSphere ESXi Image Builder 35 Required Information for ESXi Installation 74 Installing ESXi 75 Installing ESXi Interactively 75 Installing or Upgrading Hosts by Using a Script 79 PXE Booting the ESXi Installer 95 Installing ESXi Using vSphere Auto Deploy 102 Troubleshooting vSphere Auto Deploy 191 Setting Up ESXi 198 ESXi Autoconfiguration 198 About the Direct Console ESXi Interface 198 Enable ESXi Shell and SSH Access with the Direct Console User Interface 202 Managing ESXi Remotely 203 Set the Password for the Administrator Account 203 VMware, Inc.
    [Show full text]
  • Netbooting Microsoft Windows 7 and XP
    Netbooting Microsoft Windows 7 and XP Chris Holman∗ Centre for Advanced Internet Architectures, Technical Report 130226A Swinburne University of Technology Melbourne, Australia [email protected] Abstract—This tutorial shows how to set up Microsoft and testing the Windows PE are detailed in Section VI. Windows 7 and XP to be booted using network attached Installing to the iSCSI drive is covered in Section VII. storage. iSCSI support is built in to a Windows 7 installer The report concludes with Section IX and an appendix. using the Windows 7 Automated Installation Kit. DHCP, an iSCSI server and a TFTP server are configured on the II. BACKGROUND server to provide a PXE client and an iSCSI target. Once A. DHCP set up, a client PC will be able to operate entirely without a local disk. Dynamic Host Configuration Protocol’s (DHCP) [2] Index Terms—Windows 7, Windows XP, PXE, netboot- primary purpose is to assign IP addresses and other ing network configuration details to hosts. It is used in this tutorial for the purposes of providing network and PXE I. INTRODUCTION booting configuration details, including TFTP server PXE (“Preboot eXecution Environment”) [1] booting, settings and configuration filenames. or net booting, can be used to load an operating system B. TFTP over the network without using the local hard drive. While FreeBSD and Linux Operating Systems (OSes) Trivial File Transfer Protocol (TFTP) [3] is a simple support a number of different PXE boot methods (includ- protocol to transfer files. It is commonly used in PXE ing NFS, TFTP, HTTP, iSCSI and more), Windows only scenarios because of its simplicity.
    [Show full text]
  • Cobbler Documentation Release 2.8.5
    Cobbler Documentation Release 2.8.5 Jörgen Maas Nov 19, 2020 Contents 1 About 3 1.1 Release Notes.........................................3 1.2 Distribution Support......................................5 1.3 Distribution Notes.......................................7 1.4 How We Model Things..................................... 13 2 Installation 15 2.1 Prerequisites.......................................... 15 2.2 Installing from packages.................................... 16 2.3 Installing from Source..................................... 18 2.4 Configuration Files....................................... 20 2.5 Relocating your installation.................................. 21 3 General 23 3.1 Cobbler Primitives....................................... 23 3.2 Cobbler Direct Commands................................... 54 3.3 Cobbler Settings........................................ 69 3.4 Managing Services with Cobbler............................... 90 3.5 Kickstart Templating...................................... 94 3.6 Snippets............................................ 102 3.7 Package Management and Mirroring............................. 112 3.8 File System Information.................................... 114 4 Advanced 119 4.1 Advanced Networking..................................... 119 4.2 SELinux............................................ 122 4.3 Configuration Management.................................. 123 4.4 Extending cobbler....................................... 131 4.5 Power Management...................................... 135 4.6
    [Show full text]
  • Coreboot: a Viable BIOS Replacement Marc Jones Overview
    coreboot coreboot: A viable BIOS replacement Marc Jones Overview coreboot intro coreboot project status The BIOS model The coreboot model How the FS community can help coreboot coreboot is an open source boot loader project to replace the system BIOS coreboot is not a BIOS coreboot loads payloads Why coreboot? To be Free of the proprietary system BIOS Allow complete and open use of the system DRM free Security Community support Free to fix bugs and add features Name change What happened to LinuxBIOS? was: core boot code + Linux kernel The project matured Many payload options supported became: core boot code + payload Status Active Development Community 19 committers and many more contributers US conferences coreboot symposium in Denver @ HPCW SCALE 7x European conferences 25C3 , ELC, FSCONS, OpenExpo, LinuxTag Google Summer of Code 2008 All Virtual All The Time (AVATT) Status. New silicon support contributed by AMD Family 10 (Barcelona) and K8 BGA RS690 and SB600 southbridge Stefan Reinauer Core Duo VIA and coreboot community C7 CPU CN700, VT8237R 25+ new platforms supported Status.. New payloads SeaBIOS gPXE Grub2 Libpayload Status... Work in Progress Laptop/netbook support considerations System Management Mode (SMM) V3 ± 10 platforms ported Geode support K8 and Core Duo progressing Status Summary BIOS vendor model BIOS vendors have a long standing business relationship with systems and silicon vendors Silicon vendors provide initialization code BIOS vendors provide codebase and services
    [Show full text]
  • Red Hat Enterprise Linux 7 Migration Planning Guide
    Red Hat Enterprise Linux 7 Migration Planning Guide Key differences between Red Hat Enterprise Linux 6 and Red Hat Enterprise Linux 7 Last Updated: 2021-09-21 Red Hat Enterprise Linux 7 Migration Planning Guide Key differences between Red Hat Enterprise Linux 6 and Red Hat Enterprise Linux 7 Legal Notice Copyright © 2021 Red Hat, Inc. The text of and illustrations in this document are licensed by Red Hat under a Creative Commons Attribution–Share Alike 3.0 Unported license ("CC-BY-SA"). An explanation of CC-BY-SA is available at http://creativecommons.org/licenses/by-sa/3.0/ . In accordance with CC-BY-SA, if you distribute this document or an adaptation of it, you must provide the URL for the original version. Red Hat, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section 4d of CC-BY-SA to the fullest extent permitted by applicable law. Red Hat, Red Hat Enterprise Linux, the Shadowman logo, the Red Hat logo, JBoss, OpenShift, Fedora, the Infinity logo, and RHCE are trademarks of Red Hat, Inc., registered in the United States and other countries. Linux ® is the registered trademark of Linus Torvalds in the United States and other countries. Java ® is a registered trademark of Oracle and/or its affiliates. XFS ® is a trademark of Silicon Graphics International Corp. or its subsidiaries in the United States and/or other countries. MySQL ® is a registered trademark of MySQL AB in the United States, the European Union and other countries. Node.js ® is an official trademark of Joyent.
    [Show full text]
  • Coreboot Tutorial
    Coreboot Tutorial ...as used in Chrome OS (YMMV) Presented at OSCON 2013 http://goo.gl/jsE8EE Agenda (Don't panic - not every section is of equal length) Intro / Background Chrome OS Firmware Development System Preparing a Test System Hands On What Next? Who are we? ● Ron Minnich ○ Started LinuxBIOS in 1999, been working on it since. Knows everything. Couldn't be here today. ● Stefan Reinauer ○ Renamed the project to Coreboot in 2008, been working on it since 2001. Knows almost everything. ● Bill Richardson ○ Was just going to watch, until Ron backed out. Not the former Governor of New Mexico. Hi. We work for Google, but don't speak for them. All opinions are our own. What is coreboot? ● http://www.coreboot.org ● A Free Software (GPLv2) project to replace the proprietary BIOS in modern computers. ● Originally intended for clusters, now used all over the place. ● It performs just a little bit of hardware initialization and then executes a payload. ● Lots of possible payloads: Linux kernel, GRUB2, Open Firmware, Etherboot/GPXE, SeaBIOS, ... What is Chrome OS? ● "A fast, simple, and more secure computing experience for people who spend most of their time on the web." http://www.google. com/chromeos ● Chrome OS is only available on hardware. ● But Chromium OS is the open source project, with code available to anyone. http: //www.chromium.org ● We'll take advantage of that today... Coreboot in Chrome OS ● The first three Chromebooks used a proprietary BIOS, based on UEFI. ● The newer x86-based Chromebooks use Coreboot, with U-Boot as a payload. ● ARM-based Chromebooks use only U-Boot.
    [Show full text]
  • Red Hat Satellite 6.6 Provisioning Guide
    Red Hat Satellite 6.6 Provisioning Guide A guide to provisioning physical and virtual hosts on Red Hat Satellite Servers. Last Updated: 2020-04-28 Red Hat Satellite 6.6 Provisioning Guide A guide to provisioning physical and virtual hosts on Red Hat Satellite Servers. Red Hat Satellite Documentation Team [email protected] Legal Notice Copyright © 2020 Red Hat, Inc. The text of and illustrations in this document are licensed by Red Hat under a Creative Commons Attribution–Share Alike 3.0 Unported license ("CC-BY-SA"). An explanation of CC-BY-SA is available at http://creativecommons.org/licenses/by-sa/3.0/ . In accordance with CC-BY-SA, if you distribute this document or an adaptation of it, you must provide the URL for the original version. Red Hat, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section 4d of CC-BY-SA to the fullest extent permitted by applicable law. Red Hat, Red Hat Enterprise Linux, the Shadowman logo, the Red Hat logo, JBoss, OpenShift, Fedora, the Infinity logo, and RHCE are trademarks of Red Hat, Inc., registered in the United States and other countries. Linux ® is the registered trademark of Linus Torvalds in the United States and other countries. Java ® is a registered trademark of Oracle and/or its affiliates. XFS ® is a trademark of Silicon Graphics International Corp. or its subsidiaries in the United States and/or other countries. MySQL ® is a registered trademark of MySQL AB in the United States, the European Union and other countries.
    [Show full text]
  • Red Hat Satellite 6.9 Provisioning Guide
    Red Hat Satellite 6.9 Provisioning Guide A guide to provisioning physical and virtual hosts on Red Hat Satellite Servers. Last Updated: 2021-05-12 Red Hat Satellite 6.9 Provisioning Guide A guide to provisioning physical and virtual hosts on Red Hat Satellite Servers. Red Hat Satellite Documentation Team [email protected] Legal Notice Copyright © 2021 Red Hat, Inc. The text of and illustrations in this document are licensed by Red Hat under a Creative Commons Attribution–Share Alike 3.0 Unported license ("CC-BY-SA"). An explanation of CC-BY-SA is available at http://creativecommons.org/licenses/by-sa/3.0/ . In accordance with CC-BY-SA, if you distribute this document or an adaptation of it, you must provide the URL for the original version. Red Hat, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section 4d of CC-BY-SA to the fullest extent permitted by applicable law. Red Hat, Red Hat Enterprise Linux, the Shadowman logo, the Red Hat logo, JBoss, OpenShift, Fedora, the Infinity logo, and RHCE are trademarks of Red Hat, Inc., registered in the United States and other countries. Linux ® is the registered trademark of Linus Torvalds in the United States and other countries. Java ® is a registered trademark of Oracle and/or its affiliates. XFS ® is a trademark of Silicon Graphics International Corp. or its subsidiaries in the United States and/or other countries. MySQL ® is a registered trademark of MySQL AB in the United States, the European Union and other countries.
    [Show full text]
  • SCI & OSL Technical Excellence Symposium 09
    SCI & OSL Technical Excellence Symposium 09 © 2009 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice The Syslinux Project Overview of a modern bootloader Erwan Velu ([email protected]) Bootloaders aren’t dead ! Bootloaders are running on our systems for ages Lilo & Grub are, by far, the most known/popular But it looks like they didn't evolved since the dinosaur era “ Their development is complete, not need to spend time on it” 'The Syslinux Project' has never been such active since the beginning ( 15 years ago) Many contributors joined the project since the past 2 years Why such activity ? Bootloaders aren't such sexy.... What’s The Syslinux Project ? A collection of bootloaders ISOLINUX for booting CDROMs (El Torito) SYSLINUX for booting from a FAT filesystem EXTLINUX for booting from an EXT3 filesystem PXELINUX for booting over the network (PXE) Aims at booting systems from every possible media Developed since 1994 by Hans Peter Anvin (HPA) Historical Linux Kernel Hacker Co-maintainer of the x86 code Co-webmaster of Kernel.org What’s The Syslinux Project ? 'The Syslinux Project' is made of A core layer Low level operations Memory management Written in Assembly A com32 module framework Easy-to-develop extensions Written in C Many libraries as helper libjpg, zlib, pci, dmi, ... COM32, a modular approach COM32 modules have crystallized contributions/contributors as No ASM skills required to add new features or bugfix High level API / libraries
    [Show full text]
  • Annals of the University of North Carolina Wilmington Master of Science in Computer Science and Information Systems
    Annals of the University of North Carolina Wilmington Master of Science in Computer Science and Information Systems AN IMPLEMENTATION OF A GREEN STORAGE AREA NETWORK Tyler Loftis A Capstone Project Submitted to the University of North Carolina Wilmington in Partial Fulfillment Of the Requirements for the Degree of Master of Science in Computer Science and Information Systems Department of Computer Science Department of Information Systems and Operations Management University of North Carolina Wilmington 2014 Approved by Advisory Committee ______________________________ ______________________________ Dr. Ron Vetter Dr. Ulku Clark ______________________________ Dr. Doug Kline, Chair Abstract A reliable and useful storage area network (SAN) was created using two legacy servers and low- cost commodity hardware. Care was taken to create a “reference” implementation for education purposes by using color coded cables, careful cable management, labels, and diagrams. The SAN is intended to support a scalable number of drive-less servers that will host virtual machines. The SAN servers boot from USB to Open Solaris, leaving the hard drives dedicated to SAN use. Comstar iSCSI was used to communicate between the servers, and the Zetta File System (ZFS) with RAIDZ2 was used to protect against drive failure. IOMeter performance measures indicate that the SAN performs as well as a standard direct-attached hard drive, and approaches some local SSDs. The end result is redundant, fast storage using decommissioned and commodity hardware. [2] Contents Abstract
    [Show full text]