| Introducing the Linux Terminal Server Project |
Jim McQuillan (Apr. 18, 2000)
[NOTE: At the Linux Terminal Server Project website, you can find out how to create a Linux based diskless workstation that boots from a network server. Applications include both thin clients and embedded systems. The following is an excerpt from the documentation provided; please visit the website for full details.]
The Linux Terminal Server Project
Linux makes a great platform for deploying diskless workstations that boot from a network server. We have started an open source project to create the administration tools that will make setting up a diskless workstatione easier. This document describes how to obtain the ltsp distribution and how to install it.
This project was born out of the need to solve a problem for a customer who needed a terminal that could communicate with both an IBM AS/400 and a Unix application server. It needed to run TCP/IP, it needed to be inexpensive, and it needed to be easy to maintain. As a plus, it should allow the user to browse the web, and allow them to read and send email.
We could have used PC's running windows, the software is certainly available, but the cost would have been high. Both in terms of the initial investment, and the cost of supporting and maintaining the PC's over time. We decided that a diskless workstation running the Linux kernel and X-Windows would fit nicely as a solution to the customer problem.
We didn't really invent anything new here, we searched the web and found the etherboot and netboot packages. Then we found an inexpensive ne2000 network interface card that had an eprom socket. We searched and found an affordable eprom burner, figured out what kind of eprom chips to buy, learned about bootp, xdm, nfs-root and all kinds of other things. Then, we put it all together. Initially we tried it on an old 486 PC we had lying around and it worked pretty well.
So, we installed a server and 11 workstations. The customer loved it. They ordered an additional 22 workstations, and they will probably add many more in the future. (Update: They now have 70 workstations and they have ordered 30 more!) After several months, and virtually no support problems with the workstations, we have decided to share our solution with the rest of the world.
When the workstation is booted, it does the following:- It gets it's ip address from a bootp or dhcp server
- It downloads the kernel from a tftp server
- It mounts it's root filesystem from an nfs server
- It loads the X-Server software into memory and begins executing it
- It contacts an XDM server and allows the user to log into it
In most cases, the bootp server, tftp server, nfs server and xdm server will all be the same machine, we will refer to that simply as the 'Server', and we will refer to the diskless workstation as the 'Workstation'.
Once the workstation is booted, and the user logs in, any application programs that they invoke will be running on the server, while the output is being displayed on the workstation. This is a fundamental feature of X-Windows. The workstation is only running the Linux kernel, XFree86, Init and possibly a print server daemon for printing to the local printer.
Because there is very little running on the workstation, you can get away with a fairly inexpensive, low-power machine. We found in our initial testing that an i486 with 16MB of ram was actually pretty good at being an X terminal.
We used the Etherboot package available from: The Etherboot Home Page. That takes care of the image that gets burned into the eprom, and it helps us prepare the kernel for downloading to a workstation, but it stops there.
There are several HOWTO's that explain how to setup a server for diskless booting of a single workstation, but they don't discuss the problems involved in serving many workstations from a single server. The problem is that when the workstation is running, it needs to write to some files on the server, so each workstation needs to mount it's own unique root filesystem. If you had 50 workstations, you would need 50 directory trees exported. This can be a real pain to try to manage.
We have developed a method of setting up the root filesystem hierarchy that can be shared among all of the workstations. The kernel mounts the filesystem in readonly mode, then mounts a 4MB ramdisk as it's /tmp filesystem. While the kernel and XFree86 are running, they like to update a few files, so we have placed those files on the ramdisk, and created symbolic links to them in the proper place within the hierarchy.
Additionally, we have created a configuration file and a program that runs as part of the workstation bootup sequence. Each workstation can include different hardware. Things such as Network board, Video Card and the type of mouse can be configured either as a default, or individually for each workstation.
Because there aren't any application programs running on the workstation, we don't need a swap device, although it is possible to configure the kernel to swap to an NFS filesystem.
This method of booting a workstation is being used very successfully on network with 70 workstations all running from a single server running RedHat 6.0 on a 400mhz Pentium-II. Each workstation is a 166 MHz Pentium machine with 32MB of ram. We could have used a smaller processor, but these days it is getting pretty tough to find the low-end cpus.
In the future, we would like to offer the ability to run more programs on the workstation. Many people have suggested that they would like the window manager and a web browser such as Netscape running locally on the workstation.
The above is excerpted from the documentation at ltsp.org and is Copyright 2000 James A. McQuillan. Permission to distribute and modify this document is granted under the GNU General Public License - Version 2, June 1991. For the full text of the License, you can view it at: www.ltsp.org/license.txt.
(Click here for further information)
|
|
|
FUEL Database on MontaVista Linux
Whether building a mobile handset, a car navigation system, a package tracking device, or a home entertainment console, developers need capable software systems, including an operating system, development tools, and supporting libraries, to gain maximum benefit from their hardware platform and to meet aggressive time-to-market goals.
Breaking New Ground: The Evolution of Linux Clustering
With a platform comprising a complete Linux distribution, enhanced for clustering, and tailored for HPC, Penguin Computing¿s Scyld Software provides the building blocks for organizations from enterprises to workgroups to deploy, manage, and maintain Linux clusters, regardless of their size.
Data Monitoring with NightStar LX
Unlike ordinary debuggers, NightStar LX doesn¿t leave you stranded in the dark. It¿s more than just a debugger, it¿s a whole suite of integrated diagnostic tools designed for time-critical Linux applications to reduce test time, increase productivity and lower costs. You can debug, monitor, analyze and tune with minimal intrusion, so you see real execution behavior. And that¿s positively illuminating.
Virtualizing Service Provider Networks with Vyatta
This paper highlights Vyatta's unique ability to virtualize networking functions using Vyatta's secure routing software in service provider environments.
High Availability Messaging Solution Using AXIGEN, Heartbeat and DRBD
This white paper discusses a high-availability messaging solution relying on the AXIGEN Mail Server, Heartbeat and DRBD. Solution architecture and implementation, as well as benefits of using AXIGEN for this setup are all presented in detail.
Understanding the Financial Benefits of Open Source
Will open source pay off? Open source is becoming standard within enterprises, often because of cost savings. Find out how much of a financial impact it can have on your organization. Get this methodology and calculator now, compliments of JBoss.
Embedded Hardware and OS Technology Empower PC-Based Platforms
The modern embedded computer is the jack of all trades appearing in many forms.
Data Management for Real-Time Distributed Systems
This paper provides an overview of the network-centric computing model, data distribution services, and distributed data management. It then describes how the SkyBoard integration and synchronization service, coupled with an implementation of the OMG¿s Data Distribution Service (DDS) standard, can be used to create an efficient data distribution, storage, and retrieval system.
7 Advantages of D2D Backup
For decades, tape has been the backup medium of choice. But, now, disk-to-disk (D2D) backup is gaining in favor. Learn why you should make the move in this whitepaper.
|
|
|
|
|
news feed