AWS: The Infinite Self-Healing Machine
This is intended as the first of a series of articles on the benefits of the Cloud as a self-healing prospect. The series starts at a high conceptual level and gradually focuses in on specifics.
Around about 2016, people started talking about self-healing roads. Attributed to Erik Schlangen, a Dutch civil engineer, the self-healing road uses heat induction and metal fibers to easily close gaps within the structure. It is not, one hundred percent ‘self’ healing yet but that is the direction we are going, because the benefits are clear. Self-healing roads are more durable, less expensive to maintain and more secure.
A road has an optimal operating condition. When it’s at its broadest, smoothest, flattest and coolest condition, then traffic will move along it most cheaply and efficiently. We might say that this is the optimum health of the road.
Using this concept of optimal health, we can extend the self-healing concept from machines to road. Because machines are more complex and have more moving parts, their optimal health most likely will be more complex than that of roads. But they will, nonetheless have one, and it will be the optimal operating conditions of the machine. Any specific car, for example, will have its optimal fuel, tires, gearing and so on. An ideal self-healing car would, in theory, be able to maintain all of these optimal operating conditions without human intervention. It would also have the self-healing benefits we have already mentioned, being durable, relatively cheap to maintain and safe.
Now let’s take one further conceptual step and imagine a machine for making and running other machines. That machine will have its own optimal conditions, which are largely fixed and it will also have many extra optimal operating conditions, which will be as numerous and varied as the machines it can produce. It will, in short, have lots of different ways of being healthy. Each one will depend on how any specific machine within its store of machines is being used.
‘A machine for making machines’ is, of course, one way to describe a Turing machine. It could also describe your cloud account. Your account is a machine with many uses, and consequently, many potential optimal conditions. In each case, the optimal condition will be both complex and dependent on how a particular application is used. The latter will also, more than likely, be tied to business objectives.
So the question now is whether and to what degree the AWS Cloud can be described as a self-healing prospect. AWS maintain a principle of shared responsibility. This distinguishes between events that occur ‘in the Cloud’ and events that are part ‘of the cloud’ itself. This distinction maps onto the Turing view of the cloud also. Events that are part ‘of the Cloud’ are events within the Cloud machine. Events that are ‘in the Cloud’ are events within the machines made by (or on) the machine. The latter are, essentially, the applications that you create and use in your account.
These applications are — or at least can be — highly self-healing. There are methods and services in AWS for detecting and maintaining the optimal health of your applications, in almost real-time. There are also ways to fix ‘damage’ and wear on these applications, which, as has been mentioned, is connected to that applications optimal health, which is connected to your business objectives.
So your AWS account is, in fact, an infinite self-healing machine. It can give you the benefits of durability, security and cost-effectiveness that any other self-healing entity has, but for each and every one of your applications, and consequently for your business.
I will dive deeper into these mechanisms in future articles.
