At the time of its conception, virtualisation was deemed to be the next logical step in the continuous evolution of networking. Networks would be smaller, faster and more intelligent. Physical assets would at last be used to their full potential with a more intelligent approach to networking. But has the theory really fitted in with reality? As the new technology has bedded in, traditional approaches to infrastructure operations management are simply incapable of keeping pace with the dynamic nature of today’s virtual environments.
If the era of the physical environment could be defined as one of inefficiency, how will the virtual environments of today be defined? Will we have learnt from our mistakes and moved onto highly efficient virtual environments or will we be standing at the edge of a virtual abyss dealing with similar problems? With virtualisation it was typical to see 80% of a company’s time spent identifying the root cause of a problem and 20% spent fixing it. These stats can’t be replicated in today’s virtual environments. It is vital that the virtualisation management tools of today help administrators steer clear of the pitfalls of the past.
As with most new technologies it has taken sometime getting to the point where the tools on offer now are actually answering the questions that administrators have been asking. When the large switch from physical to virtual environments really started to kick in, a number of major obstacles had to be overcome. Developers struggled to really understand the impact of virtualization, how best to manage it and how to develop ways of performing Root Cause Analysis (RCA). Once manufacturers had resolved this, the market was flooded with the first generation of virtualisation monitoring tools all claiming to be able to do RCA and be the answer to all virtual management problems. These claims were not wholly accurate.
What many of these products actually offered was Alert Suppression (alarm filtering, suppression and aggregation). Although of value, it was not RCA. In its simplistic form all these tools really did was reduce the amount of information an administrator had to deal with. Administrators still had to pinpoint issues and still spend time troubleshooting. Administrators were still struggling with an inefficient approach to overall virtual environment management and problem solving. With the onset of big data a tipping point was reached with the realization that the current tools of the trade were no longer fit for purpose. So developers stepped back and asked themselves: What exactly do administrators in this day and age actually need from todays virtual management tools?
The latest generation of analytic tools now looks at troubleshooting and RCA as redundant actions for the simple reason that there are just too many permutations within the virtual environment to make them worthwhile and cost effective virtual management tools. The ‘simple’ answer was to automate the entire operations management with only the minimum of input from administrators. These tools were built with the ability to learn the ins and outs of unique virtual environments and make suggestions and take actions in order to automatically maintain virtual environments in their most efficient state. The concept of the efficient or ‘ideal state’ reflected a realisation amongst developers and administrators that keeping the virtual environment operating as efficiently as possible was a much more proactive approach to virtualized network management. Instead of receiving alerts when problems occur, intelligent analysis and control prevents alerts from happening in the first place which in turn optimises performance, maximizes infrastructure efficiencies and reduces operational costs.
IT Administrators have the unenviable task of keeping happy large numbers of end-users who do not fully understand the complexity of managing and running today’s operating systems and applications. The simple, easy-to-use experience that most end-users now get, has left an impression amongst them that this simplicity is replicated throughout the IT industry. This assumption is unlikely to change but as applications and systems do become quicker and more complicated it becomes absolutely vital that IT departments and administrators start surrounding themselves with the right tools so that they can match these end-user expectations because there is nothing like a complete IT outage to make the following words very real: Don’t fail … everyone’s watching.