Today, we are seeing that Outlook and Teams are not working for hours. There are many angles from which this phenomenon can be observed, and one interesting angle is bounded complexity
Bounded complexity is the idea that a system should be designed so that its overall structure, dependencies and failure modes remain understandable and manageable by the people responsible for operating it.
Build systems that human beings can understand, maintain and repair.
Every local system can experience failure, but if that system is sufficiently small, people will be able to understand it, and by knowing its architecture, they will be able to pinpoint the cause of error and solve it.
But in these systems which are so big so there is not a single person who can comprehend it, pinpointing an issue is the big problem. Nobody in the company can not and do not understand that system as whole.
The question is - (ok I know the answer, but rhetorically) why are they allowed to make such critical infrastructure so complex that nobody can comprehend it? I would expect that nuclear plant MUST be created sufficiently simple so it can be understood, and why world allow big companies to create complex cloud solutions they are not able to grasp?
This is just one angle, and there are many similar angles and questions to be asked about such reliance, critical infrastructure, risk etc.
Нема коментара:
Постави коментар