Problem: Our computer labs are open from 8AM to a little after midnight. Students want to work around the clock but need access to expensive specialized software.
Problem: Many of our students study abroad, live off campus, go home on weekends sometimes, etc. When they're not on campus, they don't have access to their files or to the aforementioned specialized software.
Problem: Our faculty work extremely hard but don't want to be tied to campus. But, when they're not on campus, they face the same challenges as our students — they don't have access to our specialized software and their files.
Problem: We have only so many computer labs and can't afford to simply add them on a whim, nor do we have the space. As time goes on, more and more faculty need access to computing resources in their classrooms.
The solution: "Ubiquitous computing"
Yeah, I work in academia, so people like big words, but the general theme is anytime, anywhere computing and anytime, anywhere access to resources. In short, we're working on building an infrastructure and our services based on a person's affilitation with the college, not on their physical location. From anywhere—as long as the person has an Internet connection—someone affiliated with the college should be able to access the resources that they would were they on campus. Further, we should have the ability to create an on-demand computer lab whereby a classroom full of laptop-toting students can turn the room into a full lab with their own machines and our licensed software.
Now, I'll be the first to admit that getting to this kind of service infrastructure isn't a new concept, but it is new to Westminster and I'm excited to undertake the project. It might even be looked at as a sort of utility computing initiative. I'll break this article down into two sections: 1) How we plan to get to this point; 2) The challenges that we face.
Our solution is based around Windows Terminal Services and Microsoft Application Virtualization with Terminal Services being the primary distribution method. In order to keep costs at an affordable level, we're relying on Windows Server 2008's significantly enhanced Terminal Services - enhanced printing, RemoteApp, and a ton more. We'll also use Microsoft's Application Virtualization (formerly SoftGrid) to keep Terminal Services usable and allow users to better customize their desktop and application experiences. I will admit that we're still in the learning stage on how the pieces will fit together, but we're getting there!
The way I see this working is this: When a student wants to use a licensed application that is normally available only in a lab, he will launch it from the Terminal Server using RemoteApp, which will minimize the user's need to interact with a Terminal Services desktop. With RemoteApp, an application runs from a terminal server, but in its own application window on the user's desktop. It looks and acts like it's running locally, though. Since we're not requiring users to run a complete TS desktop, we should minimize confusion.
Now, when an instructor wants an impromptu visit to a lab, she doesn't need to worry about room availability. Her students can simply pull out their laptops, connect to the wireless network that was installed last year, and launch the necessary application... on the fly. Location and dedicated space no longer become a barrier to education.
We'll also use this method for some of our administrative enterprise applications to simplify some of our desktop maintenance. In this case, the goal will be to ease the desktop support burden, but also to allow people such as our admissions counselors secure remote access to data they need in order to do their jobs. Many of our applications will only need to be maintained at the server level rather than at each individual desktop. We're still early on in this project so the full scope of possibilities hasn't become apparent yet, but I can see definitely business continuity possibilities as well.
Getting to the final solution poses a number of challenges, including:
- Manpower: We're a small group and this project will add a lot to the environment. Over time, I expect that we'll ultimately reduce the overall support burden, but the interim period will be difficult.
- Learning curve: There's a lot to this stuff. We're working on learning how to manage Windows Server 2008's Terminal Services and SoftGrid. Both are relatively complex products.
- New to market: W2K8 is brand new as is Microsoft Application Virtualization. That said, I'm comfortable with both solutions and am confident that we'll make our way through any problems. What we're trying to do is not rocket science and has been done in a lot of other places, albeit with different products.
- Security: Obviously, we need to protect ourselves. We're working on a security plan for this infrastructure as well. we want the right people to be able to use the services, but not open ourselves up to attack.
Some will ask why I have selected the products that I have for this project. Simply put, we get our Microsoft software under a campus agreement and the price is very, very good. Until we run into something insurmountable, I'll stick with this solution.
As we work through this project, I'll report back on both the technical progress and the progress toward achieving the goal of ubiquitous computing!
Since 1994, Scott Lowe has been providing technology solutions to a variety of organizations. After spending 10 years in multiple CIO roles, Scott is now an independent consultant, blogger, author, owner of The 1610 Group, and a Senior IT Executive with CampusWorks, Inc. Scott is available for consulting, writing, and speaking engagements and can be reached at firstname.lastname@example.org.