Can Linux break into banks?

What kind of progress is Linux making in financial services, and is it anywhere near challenging Microsoft's dominance? Linux's appeal is obvious: Anyone can make changes to the source code and redistribute it. And since no one "owns" Linux, this brings obvious development and licensing cost savings to Linux users.

In financial services, Linux's greatest inroads have been made in the back office. One area where it has scored well over Windows and UNIX is in terms of pure processing power. Consider the situation at banking software company ABK Systeme, based in Germany. It had requests from users for transaction rates of 100 million per day, peaking at up to 20 million an hour.

It seemed that Unix could handle a peak of around 8 million a day. ABK tweaked the performance, and stacked servers up, but it was still struggling to deliver the new levels of demand. So it turned to Linux running on its IBM zSeries hardware. Transaction volumes increased dramatically — as verified by Forrester Research's subsidiary Giga — to in excess of 19 million an hour and 155 million a day. The company now has dozens of banks running on Linux.

The growth of Linux in the enterprise can also be attributed to the fact that it is now simply more visible to corporate users by virtue of new distribution networks. For example, at the start of the year, Novell unveiled a migration path to Linux for its Netware operating system: This is a "toe-in-the-water" offer, allowing companies to try out Linux with relative ease.

So much for success in the back office. When it comes to the front office, Linux's progress is piecemeal. For example, Reuters ported its Market Data System to Linux and in the three years since then, about 20 of its largest customers have moved to the new platform. Alternatively, Red Hat in Europe, a company that provides support and maintenance contracts for Linux, reports that financial institutions are in the top three of its revenue sources.

So, while the back office may provide clear opportunities for Linux — capitalising on performance — its leverage in the front office is more complex. "For most companies, the cost to migrate away from Microsoft Windows is simply too high and outweighs the benefits companies expect they will receive," says Michael Silver, research vice president in Gartner's Client Platforms group. "Most large companies have hundreds, if not thousands, of applications, and the cost to migrate them to run on or be accessible from Linux clients is huge."

Another way of assessing the current state of play is to examine the battle between Linux proponents and Microsoft executives over which is best. "Pros" on one side are routinely reinterpreted as "cons" by the other. For example, Linux advocates say that it has a lower total cost of ownership (TCO). They point to the fact that you don’t have to pay for a new licence every time you deploy a new copy of the software. Moreover, the bureaucracy associated with managing these multiple licences also represents a significant cost — one not borne by Linux users.

On the other hand, as Microsoft points out, TCO does not depend only upon start-up costs. If a system crashes, or needs upgrading, or is wanted to do something different — this all costs too. Windows is arguably a cheaper option on that front. Another factor that is routinely thrown into the ring is security. The prevalence of Windows-oriented viruses is obvious, as is their ability to spread fast. But then, Microsoft is throwing billions at the problem, and at the end of the day, simply switching the operating system will not make you safe from hackers or phishing scams.

Perhaps the truth of the matter is that banks will switch to Linux when, and only when, it makes financial and technological sense. These days, there are few ideologues in IT departments. So, Linux must make its case to continue gaining ground. But then, so must Microsoft, if it is not to lose its current, substantial lead.