During the start of my career as a Software Engineer, my first assignment was to maintain a COBOL system that used to transact approximately 5000 records per hour. It was very huge and challenging system with web and AS 400 system integration. During the start of the career, the general idea for me as a Computer Engineering student was to build software framework and systems with fancy programming languages and databases. Once i was thrown in to the COBOL maintenance, i was kind of dejected for initial few weeks. Luckily, my manager noticed this and made me understand why is it important to maintain software systems and what can one learn from it.

I am writing this article to remind all the developers and designers of the software/hardware systems in all industries about the maintenance of the critical systems. A problem, which everyone thinks small could become big or crucial or critical in certain circumstances. All the industries face the same problem that any system can not be tested with all the real time scenarios. The test data or test cases for any system are limited and time bound, So can not be trusted for 100% test coverage and safety of system.

Very often we encounter the “refactoring” dilemma in the software industry. The question that comes to everyone’s mind is should we refactor “NOW” or put it off for later “trigger” ? All projects are faced with the following challenges, which makes a project to decide if a “refactoring” is necessary at that time.

• short time
• limited budget
• non-availability of resources
• pressure from sales and marketing and
• finally pressure to deliver

We always tend to postpone and procrastinate the code, design and architecture refactoring. Very often “shit happens” and the cost of refactoring is sky rocketing. Customer is angry, development team gets demotivated and project stakeholders are unhappy with the system performance. Some of these problems are addressed by the agile methodology (TDD, SCRUM, XP, RUP, etc…) and some are addressed by the timely act of “experienced” leaders in the industry. However good is a methodology or a process, finally everything depends on the people who implement it. So many times i get “upset” when big organizations talk about “people independant” process ???

Finally, i was moved by the recent incident of the Air France flight (Rio de Janeiro to Paris) havoc, which probably seems to be a problem with some failed hardware. The news seems to be that the hardware sensors had to be replaced some months back and for some reason they did not do it. Irrespective of whether this is a hardware failure, it calls for everyone to be more attentive, proactive  and creative when building the critical applications and systems. Following is an excerpt of the news from internet.

Air France issued a statement with details about the monitors hours after the French agency investigating the disaster of Flight 447 said the instruments were not replaced on that aircraft – an A330 – before it crashed last week into the Atlantic Ocean en route from Air France issued a statement with details about the monitors hours after the French agency investigating the disaster of Flight 447 said the instruments were not replaced on that aircraft – an A330 – before it crashed last week into the Atlantic Ocean en route from Rio de Janeiro to Paris.

Air France said it began replacing the monitors on the Airbus A330 model on April 27 after an improved version became available.

Pitot tubes, located on the exterior of the aircraft, are used to help measure aerodynamic speed.

Aviation officials have said the crash investigation is increasingly focused on whether external instruments may have iced over, confusing speed sensors and possibly leading computers to set the plane’s speed too fast or slow – a potentially deadly mistake in severe turbulence.

An Air France statement said that icing of the monitors at high altitude has led at times to loss of needed flying information.

However, the Air France statement stressed the recommendation to change the monitor “allows the operator full freedom to totally, partially or not at all apply it.” When safety is at issue the aircraft maker issues, rather than a recommendation, a mandatory service bulletin followed up by an airworthiness directive..

Air France said it began replacing the monitors on the Airbus A330 model on April 27 after an improved version became available.

Pitot tubes, located on the exterior of the aircraft, are used to help measure aerodynamic speed.

Aviation officials have said the crash investigation is increasingly focused on whether external instruments may have iced over, confusing speed sensors and possibly leading computers to set the plane’s speed too fast or slow – a potentially deadly mistake in severe turbulence.

An Air France statement said that icing of the monitors at high altitude has led at times to loss of needed flying information.

However, the Air France statement stressed the recommendation to change the monitor “allows the operator full freedom to totally, partially or not at all apply it.” When safety is at issue the aircraft maker issues, rather than a recommendation, a mandatory service bulletin followed up by an airworthiness directive.