I have been meaning to start a series about a new project I’m working on. It doesn’t even have a name yet, but I have had positive results so far. This post introduces the busuness and technical drivers used during the early stages. Hopefully, I will find the time to write a full white paper on the topic this summer.
I was looking to broaden the CRT project base so I thought I would look at the systems that drive Real Estate offices. What would be nice was a project dedicated to radical cost reduction with moderate functionality improvements over current approaches. A tall order.
While talking with several of you on unrelated issues, I began reminiscing to the old days of the mainframes and "minis". Boy, weren’t things like security, backups and offering new applications simpler in those days? There was only once copy of the application and all data was in a central location. Sure, the architecture had its drawbacks, but at least everything was in one place. Today’s computing environment is a radical departure from days past. Every workstation is capable of local storage making the backup and security issues more complex. Fortunately, there are vendor solutions to manage this structure though. On the flip side of the argument are the cost and risk factors. If you choose to avoid the costs of these solutions, you risk a lot if your infrastructure fails. Installing and maintaining these solutions does not provide an economic, competitive edge either. I am seeking radical, not marginal cost reduction.
Some of the new mobile devices embrace the terminal thoughts of the past. The phones are actually control devices for centralized services. Your e-mail, contact lists and calendar are actually not stored on the device. If the device is damaged, your address book is not lost. I thought the approach taken by mobile devices might be able to be replicated with desktop computers. Could I make desktops operate local devices like sound cards and video cards yet still rely on centralized services. I wanted to explore this as an alternative to distributed backup, maintenance and security packages. Some would consider this thinking alone as radical in the technical sense.
Skipping forward in this intellectual debate a bit, I cracked open some fully depreciated computers to look at what kind of device could be built from "free" (financially speaking) equipment. What I found was not shocking. Old CPUs and minimal memory. There were some promising finds as well. Most of these computers had AGP video cards, modern sound cards and adequate network cards. Good. I have modern devices with slow CPUs and memory constraints, making the project worth pursuing. Mobile devices also have limited CPU and memory resources but manage to run perphiral devices like keyboards, sound cards and cameras.
The first thing I need is an operating system that has a (very) small footprint. Even underpowered CPUs can effectively drive the devices, but I have to be memory conscious. Because my plan is to remove the hard drives completely, being able to load the entire OS from a CD-ROM or over the network is a requirement. If any of this sounds familiar to the thin client concept you are right. I’m not re-inventing anything here. I just wanted to make sure that I arrived at an answer based upon sound reason versus picking a technology and making it fit the problem.
Now I would like to begin adding some operational and technical "wish list" items. I would like to support both Windows and Linux based terminals, so that the users can choose the environment they prefer. Choice is good; any user should be able to run the desktop of their choice. I would also like to authenticate all users with a common mechanism (OpenLDAP or ActiveDirectory) so that the solution can fit into existing operations. Part of this approach would be to create account structures (home directories) on the central server for the user if they have an account, but have never logged into the server. Users should be able to log in from any terminal have access to the sound card for their applications.
Now for the big one, financial considerations. How can I achieve all of this with a minimal financial investment? Well, this is where Part II will begin.
0 Responses to “What was old may actually be new - Part I”