“Computers are incredibly fast, accurate and stupid; humans are incredibly slow, inaccurate and brilliant; together they are powerful beyond imagination” is a famous quote from Leo Cherne. And indeed, it has made sense to delegate all those mundane tasks that required nothing more than raw power, mindless speed and uncompromising efficiency to computers. Freed from the mundane, businesses could now focus on higher ideals like strategy, growth and innovation. It was division of labour made in heaven.

The evolution of technology and its promise of automating increasingly complex processes drove enterprises to stockpile IT. Eventually, that snowballed into underutilised storage resources, unproductive data centres, a superabundance of strategic vendors and an antiquated approach to maintenance and support. And pretty soon enterprise IT teams were back to doing the mundane - maintaining and troubleshooting an intricate network of heterogeneous IT systems. With the result that today, the promise of automation is ironically contingent on manual intervention. 

Regression apart, this severely hamstrings enterprise abilities to pursue future growth imperatives. With up to 80 percent of IT resources consigned to bringing current systems up to speed, enterprises have only the meager leftovers to invest in innovation or emerging business transformation opportunities.

Certainly, they can modernise infrastructure and adopt integrated, optimised sourcing strategies to restore IT systems to enterprise grade performance standards. But for a truly transformative impact on efficiency, they will have to turn to autonomics.

Automation is the deployment of systems, rather than humans, to perform certain tasks. Autonomics automates mundane, repetitive tasks in the operation and maintenance of said systems. Inspired by the human autonomic nervous system, the most sophisticated, naturally existing autonomic system known to man, autonomic computing represents the next evolutionary level in automation. It is automation automated, so to speak.

Take the example of batch operations on a network. These processes are generally run at night when there are no users, activated by a piece of code written for this purpose. But a human element is still needed to ensure that the code works and is initiated at the right time. With autonomics, the system itself becomes responsible for scheduling and executing the operation, with no manual intervention whatsoever.
           
Autonomics is based on expert systems that, unlike script-based automation tools, enable self-learning and self-healing. Over a period of time, these systems can ‘learn’ an IT environment, identify and resolve issues, and independently execute a range of tasks. In the context of operation and maintenance of IT systems, autonomics eliminates manual intervention in routine tasks like information gathering and troubleshooting, tasks that currently require skilled IT personnel.

That being said, the case for autonomic computing is built on realities that are much bigger than the squandering of skill on the ordinary. The sheer pace of proliferation of computing devices - 2.4 billion computers, tablets and cellphones will ship this year alone - is outstripping the supply of skilled resources required for support and maintenance. Even assuming we find the manpower, the complexity of these systems will stretch the limits of human ability. Because to ensure sustained system performance in an always-on world, decisions have to be made on the fly and issues resolved as they happen - clearly capabilities currently beyond the scope of human construct.      

This is precisely the impasse that the human construct-inspired autonomic computing systems can resolve by bringing four significant enhancements to any IT infrastructure.

  • An ability to self-configure and dynamically adapt to changing IT environments in real time.  For instance, they can automatically respond to increases in Internet traffic by reassigning application servers as Web servers, replete with the software required for the temporary role. They can also reset servers to their original configurations once traffic has normalised. 

  • An ability to self-optimise resource allocation and utilisation to maximise operational efficiency and performance. They are also capable of learning to optimise in the context of a stated enterprise objective. For instance, to achieve the objective of completing a complex task within a stipulated time frame, autonomic systems will optimise between available resources and competing tasks to identify the best way to get the job done.

  • Self-healing capability to detect or even anticipate extraordinary events, like component failures, and take corrective action to prevent system disruptions. Even if some of their parts are compromised, these systems can reroute workloads while they attempt to fix the problem. Most importantly, autonomic systems learn from their experiences to inform future decisions and prevent recurrences.

  • A self-protecting feature, which identifies malicious behaviour and autonomously activates protective measures.

Through all this, the only direct role for human beings is to define the overall policies and rules that guide these self-managed systems.    

Over the years, IT maintenance and support capabilities have not been able to match the pace of growth and adoption of enterprise technologies. Today the lag between the two is so marked that it threatens to undermine the fundamental promise of IT. Before enterprises move to the next level of disruptive technologies, they need to disrupt the way maintenance and support services are delivered within the current technological framework. Autonomics is one way to do so.

Chandrashekar Kakal is Senior Vice President and Global Head of Infosys’ Business IT Services
Enhanced by Zemanta

Find your next job with techworld jobs