Like a big portion of the world’s populations, I very much enjoyed the recently concluded 2012 London Olympics. Over 10,000 athletes from 204 countries competed in 300 events. The Summer Olympics are just about the biggest possible stage for everyone involved, - the athletes; the host cities; and the different companies involved in the production of the Games. For each of them, it is an opportunity to shine, or to stumble, while the whole world is watching.
The Olympics are an excellent occasion for companies to showcase their capabilities in a very demanding environment. So, having recently launched IBM’s Internet initiative, we were looking forward to unveiling the very first Olympics website at the upcoming 1996 Summer Games in Atlanta, as part of our efforts to establish IBM as a major Internet player.
We were pretty sure that the Atlanta Olympics website was one of the largest Web project anyone had undertaken so far. We were well aware of the considerable risks inherent in doing such a complex, new project on such a global stage. We were worried that beyond a certain number of users, the response time would start to degrade, and if sufficiently stressed beyond its capabilities, the system could become unstable and crash.
But, we felt that this was exactly the kind of market experiment we should be doing to learn what was needed to support large numbers of live users on the Web. We were part of a wave of innovative, experimental projects that were trying out all kinds of new Web applications. We were counting on our Internet audience to understand that this was not yet a prime time application - more of an alpha or beta system, - and thus forgive its shortfalls. We got as ready as we possibly could before the games opened - and then crossed our fingers hoping that things would go well.
IBM had been the official technology sponsor of the Olympics since 1960. The Olympics were viewed as a unique opportunity to demonstrate IBM’s technical prowess in front of customers and the world at large. In the early years, IBM’s key contribution was the main computer system that captured the results at all sporting venues and distributed them to scoreboards, judges, commentators and others within the Olympics site. Over the years, more sophisticated technologies and applications were added.
For the Atlanta Games, IBM was not only the major technology provider, but also acted as the lead solutions integrator, making sure that all the various systems worked together including those developed by other vendors. Among the new solutions deployed in Atlanta were an information system with kiosks spread throughout the Olympics site that anyone could use to learn about each of the participating athletes, the different events, schedules, transportation and so on; a system for distributing real time results and event information to news agencies, press, and radio and TV reporters around the world; and the aforementioned first official Olympics website, offering near-instantaneous results for all events, in addition to photos, news, athletes’ bios and other information.
The overall planning and development for the Atlanta Games IT infrastructure was handled by the IBM services organization, which had lots of experience given IBM’s involvement with the Olympics over the past 36 years. The website was not part of the original Atlanta planning, since the World Wide Web barely existed outside the research community when the planning started several years earlier. But, by 1995 the situation had changed. The Internet and Web were picking up steam at a very rapid pace, and the decision was made was to build the first official Olympics website.
Because this was all so very new and experimental, the responsibility for developing the Olympics web site was given to a small IBM team of Internet experts. The team would follow its own separate development cycle, more suitable to that of an experimental project than a production system. It was considered a learning experience for all involved.
The Atlanta Games took place less than a year after Netscape went public, which many consider the key event marking the Internet transition from a research network used primarily by the technical community to the commercial behemoth that it went on to become. The new capabilities made possible by the Web had the feeling of magic, but, in 1996, it was pretty primitive, slow magic. To begin with, the majority of people were accessing the Internet over slow dial-up modems, with bandwidths of around 32 kilobits per second or less. Only at work, if you were lucky, did you have access to faster broadband speeds. It wasn't until years later that broadband usage in the home became commonplace.
We used IBM's SP family of parallel supercomputers which we were confident would provide us with all the computing power we could want. But the software for web servers was quite immature. The team used the open source Apache Server as the basic web server, and custom built the extensions needed to support its content, applications and other capabilities.
For the regular production Olympics systems, the requirements and capacities were pretty well understood. But, it was not clear how many people would actually visit the website, what features they would use and how long they would stay connected. Our plan was to start out with a limited set of functions to make sure that the system worked well and could handle the traffic. Assuming things went OK, we would then add additional functions as the Games went on.
Our Olympics website worked quite well, except for some unduly slow response times when traffic got very heavy. We learned a lot about the requirements for building and operating large, complex websites. All in all, it was a very successful experiment.
But, while the experience with the website was quite positive, IBM's Olympics high wire act ran into unexpected troubles with one of the production applications. Most of the computer systems worked flawlessly, including the applications that delivered the results to scoreboards and judges. However, there were serious problems with the application that delivered the event results to news organizations and reporters around the world. These caused the results to be delivered late and in some cases they included erroneous information.
The problems were corrected within a few days, but by then the public relations damage was done. Reporters were justifiably unhappy with not being able to get the results from the various events in a timely way and wrote some pretty negative stories about IBM. It was quite a humbling experience.
IBM went on to be the lead technology sponsor for only two more Olympics - the 1998 Winter Games in Nagano Japan, and the 2000 Sydney Summer Games. The company significantly tightened the project management, adding more time for testing, doing everything possible to reduce risks, and appointing a senior executive to oversee the whole effort and report back regularly to the top of the business. The IT systems for Nagano and Sydney were flawless.
By 1998, the Olympics website was no longer an experimental system, but just another major production application, albeit one that kept growing in importance matching the rapid expansion of the Internet and its increasing impact on all aspects of society. IBM continued to be the lead integrator in Nagano and Sydney, a role that proved increasingly complex and expensive given the growing penetration of IT into just about all aspect of the Games. The International Olympics Committee felt that no one company should be the overall technology provider and the Internet should become its own separate category with different sponsors. For a variety of reasons IBM and the IOC could not come to agreement and parted ways after forty years.
As I was watching the 2012 London Olympics and appreciating all it took to put on such an event, bittersweet memories of the 1996 Atlanta Games came to mind. On the one hand, it is very painful to stumble in prime time, as anyone who has gone through the experience will attest. It is hard enough to find a totally unexpected problem in a complex, software-intensive system when things are quiet and you have the time to do so. It is something else again to do so in the middle of a crisis, with angry people around you and headlines like IT Disaster at the Atlanta Olympics or IBM’s Olympics Fiasco.
On the other hand, it is heartening to see the central role that the Internet now plays in the overall Olympics experience, - with all kinds of information, pictures and videos delivered to a variety of devices. And, it is nice to have the memories of having been involved with the team that built the very first Olympics website back at the 1996 Atlanta Olympics.