Linux: Windows Beware

By Drew Robb

(Back to article)

If anyone in the IT field deserves to say he has "been there, done that" it is Kirk Rheinlander.

You would be hard pressed to find an area of IT, or of the country, he hasn't worked in. He received his B.S. in Industrial Science from the University of Southern Maine on the North Atlantic coast. Later he did MBA work at Pepperdine University which overlooks the beach in Malibu, Calif.

On top of that, he has IBM AIX certification in system administration, Novell CNE (Certified Network Engineer), IBM management and marketing training as well as certificates in many other platforms and applications.

His professional career includes being senior engineering manager for Storage Technology Corp. and enterprise architect for Cardinal Health, a $50 billion pharmaceutical distribution firm.

But his major work for the past decade has been consulting for large corporations and government agencies.

Among other tasks, he participated in developing Amoco's I/T architecture and then established and managed the IT infrastructure to support Amoco's merger with BP, one of the world's largest corporate mergers.

As you could imagine, Rheinlander has also worked with many platforms. He was part of the team at IBM which developed the world's first PC, but he also founded the world's oldest Mac Users Group in Conejo Valley, California.

A few years ago, he was making the pages or Wired and Linux Today for setting up a Linux network at the charter school his children attended in Ft. Collins, Colo. He can't tell us how that is working out, however, because he (and his kids) are now in Dublin, Ohio, where he is principal of the small consulting firm, KPJ Squared.

CIO Update caught with Rheinlander there to see what he has been up to and where, in his view, IT, and particularly Linux, is heading these days.

CIO Update: What are you up to these days?

Rheinlander: Mostly strategic IT and business consulting, building out the long term IT infrastructure. It is the same type of work I did before, but I'm spending more time doing it from a business strategy perspective rather than a pure technology basis.

Have you seen a shift from an IT-centric to a business-centric viewpoint in your customers?

We've been preaching that for many years, but I don't know that there is a shift going on. Some companies are and some aren't, but I'm being more selective about what I do now.

I am trying to stick more to companies that will do it. Designing an IT architecture is good, but if the company never ties it to their business strategy, they never follow through with implementing it.

For me personally there has been a philosophical change. I will only deal with companies that going to tie IT in with their business strategy so that IT is enabling the business effectively.

What operating systems have you worked with in the past?

I don't know if I can name one I haven't worked with. Oh, wait, I haven't worked with Pick.

What has been your experience with Linux?

I have used it extensively in server projects and some desktop projects. The Linux terminal server project is phenomenal; one of the slickest things going.

How does Linux compare with other OSs?

It is very stable, but the human interface experience is fragmented. That causes a significant problem in getting Linux desktop adoption. The difference in system management between Red Hat and SuSE and other Linuxes out there is phenomenal in how you deal with stuff. It is frustrating.

For example, SuperDrive support is available with some versions, but it is not readily available with Red Hat. You have to go through a lot of kludge to do it and as a result you can't mount DOS based file systems easily. There are a thousand examples like that.

What about Linux on servers?

They are rock solid. When you measure mean time between reboots in months instead of hours, that is a significant advantage. When you look at the marketing hype about Windows servers they say that Windows is more stable. That may be a true statement if you don't ever install anything outside of Windows.

But, as soon as you install something else, because of the fundamental nature of how windows works and the way it is architected, it has a tendency to grow unstable with more and more applications. Linux does not have that problem.Is Linux mature enough for the enterprise?

Absolutely, especially Linux on the mainframe. That for me is the prime opportunity for servers. There is a spectacular opportunity to take large scale integration efforts.

In what cases would you advise your clients to use Linux?

Whenever possible. The only time I wouldn't recommend that a client head down that path is if they have a very strong application portfolio and/or skills mix portfolio that is very focused on another platform.

For example, I have a client that is very zOS focused. Mainframe folks. They also have a Windows contingent but have very little UNIX skills. It would be a significant cultural shift for them to deal with switching to Linux both from a skills as well as an applications standpoint.

In that case, Linux on the mainframe would be a good transition path for them since they already have mainframe skills. They have a proliferation of servers and a significant portion of them would do better in a Linux environment, but without the necessary skills it would be a difficult transition.

But the ability to put Linux on the mainframe still allows them to use all the mainframe management capabilities so they can deal with it more effectively.

When would you advise that they don't use Linux?

If they are already highly committed to an extensive application portfolio or skills. It is often an expensive transition moving to Linux.

One thing people always make a mistake with when looking at platform transition is they only look at the cost of the server box or the server operating system. But you also have your backup software, disaster recovery software, system management tools, your skill sets, your databases and what they run on.

It is the total cost of deploying a solution, not just a platform. If the cost of moving the solution does not make sense, there is no logical business reason to do it.

There is no such thing as technology for technology's sake any more. It has to make business sense.

What still needs to occur to with Linux so we see broader adoption?

A consistent system management tools and a wider portfolio of enterprise-class application support. On the desktop we need a consistent user experience and more applications, in particular an office suite. The StarOffice/OpenOffice suite is getting there, but it is not up to the level of Microsoft Office yet.

Do you see Linux as replacing UNIX, or working side by side?

I think Linux will replace UNIX eventually. The UNIX market has always been so fragmented between all the implementations. Linux, because of binary compatibility, has a better chance of adoption

What do you see as the future of Linux?

I believe it is going to become Microsoft's worst nightmare. As it matures, as it becomes a consistent human experience for the platform, we will see widespread adoption. It is too scary for most people right now, too much of an unknown, too much of a geek tool. It is not mainstream yet.

When will it be mainstream?

Twenty-four months maybe. It depends what happens with Novell and their work with SuSE, Ximian and putting that all together.