Archive for the ‘scale-out architecture’ Category

Aristotle and the Soul of the Internet

November 21, 2006

Aristotle determines what an organic living object is using the following criteria:

a. Growth, nutrition, (reproduction)
b. Autonomous Locomotion (i.e. auto-generated movement in space)
c. Perception
d. Intellect (= thought)

This gives us three corresponding degrees of soul:
a. Nutritive soul (plants)
b. Sensitive soul (all animals)
c. Rational soul (human beings)

Now, many are speaking about the Internet becoming, sooner or later, an organic creature, and so what I’d like to show is why the Aristotelian criteria reinforce and probably explain these feelings, given the advent of utility computing and the architecture titled – scale-out.

Scale-Out Architecture

Scale-out is an architecture used more and more in those cases where computer systems are required to scale ad-infinitum in order to sustain an unknown yet massive amount of online users. Amazon, eBay, Google etc. are all companies specializing in both the deployment and the optimization of scale-out architecture.

This architecture has the following conceptual components and processes:

The Brain (or mind, or manager, or controller)

The brain constantly senses (or monitors) the system’s environment: how many users are currently on-line? What is the overall CPU consumption of the system? What is the status of each of the hardware/software components that makes the system? and so forth.

The Brain performs real-time compilation of all this sensual data and meditates about the current state of things. In case a faulty situation is either identified or anticipated, the Brain reacts by adapting itself to the newly created situation. To better visualize this adaptation, I will use the following scenario:

A system consists of 4 servers. The Brain identifies a dangerous increase in users’ load – something that can be solved by adding a fifth server. The brain then launches dynamic, self-healing/self-nutrition processes that take a bare metal – a hardware-only box – attaches it to storage and network devices, installs the required operating system and applications and finally makes the server fully operational.

We have just witnessed two Aristotelian criteria in motion:

Nutrition, Self-feeding and reproduction

The system just “ate” a bare metal, digested it and turned it into something it needs. In many cases, what the system is actually doing is cloning itself into the new server – a reproduction process.

Self-induced locomotion – movement in space

The system now occupies five servers. It occupies more physical space than it has occupied a minute before.

Needless to say, all these operations occur without any human intervention. They are completely autonomous.

The Internet, hence, has a soul.

Q.E.D

Advertisements