This article gives us an in depth look at another type of IT debt: architectural debt. It starts off with the jarring statistic that 72% of IT budgets are usually spent “keeping the lights on” or in other words day-to-day maintenance. The only way to reduce this proportion of the budget dedicated to maintenance is to address its cause.
Here is where architectural debt comes into the picture. The more we move to software defined architectures, that rely heavily on integration and automation, complexity arises in the face of changes to support loosely coupled and highly integrated data center architecture. In simpler terms the implications of a a generation of architectural decisions accumulate debt as we try to move forward to more automated efficient network processes.
Example: The choice of scripting languages used has much more of an impact than is usually thought. If you have a disparate set of object models and interfaces through which services are provisioned system entropy will occur. System entropy is when “as a system is modified, its disorder, or entropy, always increases”.
Ultimately to avoid entropy, architectural debt must be reduced. Software defined architectures become a better mode through which to manage architectural debt and reduce day-to-day expenses, as they can contain architectural debt as well.
To read the full article and get a fuller look into architectural debt click here: http://devops.sys-con.com/node/3110908