I finally find the time to write to you on the integration of automated IT operating into today´s working environment. One should think that automation means that just some other tool will be installed into the void of the IT service management jungle and maybe some administrators use this tool and become a lot better and a lot faster. Actually that is exactly what I am NOT talking about. If you are interested in my opinion on the “Automation Market” you will read an article here soon. So what I am talking about is more of an auto pilot – a machine that actually looks at problems and chooses to take action.
The question I am always beeing asked is ‘does this integrate with an established working environment and established processes e.g. ITIL?’ Yes it does. Indeed it needs established processes to function properly. An automation engine like ours (ff0000;">aAE) basically replaces the initial contact to IT experts.
Let us look at the ITIL V2 and V3 incident management process for example. As you probably know (see figure 1) a normal ITIL incident management process is either initiated by an alarm from some monitoring system or by a user contacting the helpdesk. The helpdesk handles all the bureaucracy and then passes the incident on to the IT experts who will perform further analysis – if required collect additional data and perform additional analysis – and then either take immediate action to produce a solution or initiate a change process to take this action. As you can see in figure 2 in an automated incident management process the automation engine takes the place of these IT experts. This also includes the engine communicating with the helpdesk, performing additional analysis, requesting additional data, documenting its actions and so forth. When the automation engine cannot find a solution it will contact the IT experts and ask them to step in. Only this time the experts will get a well analyzed incident with most of the boring work and analysis already done and well documented so they can actually work on something new and interesting.
This is how we introduced this auto pilot into our own ITIL compliant IT service management unit. We promised the real technical experts that they would never be bored to death by everyday tasks and tedious busywork. Instead the engine puts only these problems on their desk where an expert as such is actually required and can use his or her talent instead of just keeping mindlessly occupied. If you want to read some more on the human element and concerns connected with the introduction of automation you might want to look at the article “Plays Well with Others” written by Ellen Fussell Policastro last August. In this Article automation is looked upon not in an IT sense but in an industrial sense. This environment deals with change more practically than just IT and therefore it is probably an early adopter for the automation change on its way now.
So you can see that in an environment with well defined processes it is very easy to place an automation engine or an IT operating auto pilot. In an organization that does not have IT operating processes in place yet, just finding the proper interfaces for the automation engine and redefining the roles of the IT experts is probably a piece of work for Sisyphus.
Incident management is just one example of how an automation engine that actually acts like an auto pilot can be integrated to dramatically reduce cost in IT operation while simultaneously increasing quality and making the jobs of IT experts much more interesting. As can be seen from figure 3 the automation engine places itself between CMDB with enterprise monitoring system and the process layer actually involving IT experts. This is not only valid for the reactive ITIL processes like incident or problem management but also for proactive processes such as availability or capacity management where our autopilot engine will itself invoke work load automation tools in order to up- or downscale an IT environment according to predicted usage and demand.
This level of integration into established processes and behavioral patterns of technical advanced staff is very rare for a tool that radically changes the workload of IT operating teams, service managers. So this approach is one of the few roads available to actually move one step ahead in an environment that produces ever more complex IT applications, interdependencies between IT services and speed of change within the environment. Just think about the kind of pressure a fully cloud computing based banking data center would put on administrators…
They would have to cope with a dynamically changing environment, changing dependencies and rapidly changing communication matrixes. Automation could handle the ordinary tasks in such an environment without being pressurized by speed and changing preconditions and contact IT experts with exact and well documented information when an unknown issue occurs – relieving them of the pressure generated by the dynamic IT environment and making much more use of their actual expertise. If that is not what we want (I could not really see a reason) it is certainly what we need to keep up with a changing world without further demolishing the image of the IT-Crowd.