I had a meeting with a customer the other day which was centered on virtualization and private cloud and a funny thing happened: it morphed into a Business Service Management (BSM) discussion.  The discussion got me trying to put my arms around “BSM in a virtual environment” and what it means.

In a traditional BSM scenario, you are managing your IT from a business perspective with the ability to drill into the business models until you can isolate and correlate the technology supporting the business service.  But in a virtual environment, you’ve abstracted another layer, right?  For example, my order processing is running slow.  Which is of more value to me?  The first layer down tells me that the virtual database server is experiencing a performance hit or knowing that the underlying network-attached storage is living in a network segment that is currently overloaded due to end of the month processing?

To be honest, I’m not sure there is a clear answer.  I think if I am the “break/fix” guy, I need to know at the lowest level so I can attack the problem in the infrastructure.  And of course, the business unit wants to know at the highest level that order processing is experiencing a slowdown.

What my conclusion was is that at the highest level, BSM hasn’t changed.  You are still mapping out your business processes to show the health of the process flow and the performance of the service.  But, to accurately or perhaps more appropo, usefully, map the technology to the business process, you need to have the virtual abstraction layer, which has the actual infrastructure mapped to it.  This could lead to some interesting analysis, such as response times across the virtual environment compared to response times in the physical environment and how they both correlate to performance of the business service.  Just some food for thought.

I’d be interested to hear about some real-life examples.  How are you monitoring the virtual?