 OK, we all know most organizations were over-provisioned in the 90's. The press has been full of stories of businesses not knowing how many servers they had, where they all were and, even if they could find them, they were seriously under-utilized. Hardware was cheap, business was booming, over-provisioning was the way we all mitigated the risk of performance problems. Didn't your move to virtualization correct all that? Possibly not - Gartner has been warning of 'virtual server sprawl' for some time. It's easier to roll out a virtual machine than it ever was to roll out a physical server. Virtual machines are proliferating at a faster rate than physical machines ever did, and we have still only virtualized the 'easy' candidates. This brings both opportunity and threat to the business.
The threat is that costs will again spiral out of control, driven by the easy to implement, low unit cost, incremental expansion of infrastructure that virtualization supports. The opportunity is to get the environment under control early this time around, sweat the assets, plan what we need to keep the business happy in the future and put a smile back on the CFO's face. What needs to happen is for the business requirements to drive the technology, not have the technology run ahead of the business need. Capacity management brings together the technology (what resources we need and when we need them) with the business (what we need to achieve and what it should cost). When planning for the future, the business should be the starting point.
To maintain control of your virtual infrastructure as it expands rapidly, as it surely will, means you will still need the technical aspects of capacity and performance management. Automated means of identifying under-utilized virtual resources, moving resources to where they can support requirements and similar activities will all be needed on a day to day basis. The speed of expansion of virtualization and its implementation will possibly keep increasing for some time. There is still much more of our environment that will benefit from virtualization, moves to Cloud are largely based on virtualization as the enabling technology and as we emerge from recession to growth, demands on IT will increase. If this is to be managed well and cost effectively, and all the over-provisioning mistakes of the recent past are not to be repeated, proactive planning will be needed alongside the day to day management of performance.
For some organizations, there is already some disillusionment with virtualization. Initial virtual migration was sold on promises of 25:1 virtualization ratios, 25 physical servers becoming one virtual server. Reality is averaging more in the region of 10:1. Expected savings have not been as good as anticipated. With this one must remember that the industry on average is only 30% virtualized. Future virtualization of physical systems could involve increased migration cost. Once virtualized, such systems are proving more costly to maintain than expected. True, fewer staff manage more systems. On the flip side, more fluid architectures mean change control is often weak, and unplanned change often means increased costs. Capacity management offers the tools and the processes to get this process of change under control, whether migrating physical to virtual or managing your virtual estate.
Virtualization has rapidly become too important to business success to leave to chance. As a minimum any organization needs a clear window onto what is happening on a day by day basis. In such a volatile environment, staff need to have access 24x7 to automatically updated information on where problems are, what options are available to resolve them and help in making those changes. It is not enough though. A good sailor will tell you that this is 'steering by your wake', looking backwards to see what direction the boat is going. At the same time the relevant eyes need to be looking forward. Those eyes need to be fixed on the business horizon as well, not just on the boat itself. Where the business needs you to be governs day to day decisions you are taking now. Capacity management is the link between the technical data needed to control the boat and the navigational knowledge of what direction you need to take.
Business and technical views need to come together to manage infrastructure effectively. For any large organization, that infrastructure is not just virtual. It encompasses an ever more complex array of physical and virtual environments, now meshed with a variety of Cloud services. In the past capacity decisions have often been taken in isolation by IT management. The increased variety and flexibility of processing options, both within the business and without, means that this can no longer be the case. Capacity decisions will be made at a more strategic level, and must be made from a more 'real' business perspective. What the business needs to achieve needs to be mapped onto the best technological option available to achieve it. Whilst the answer might often be 'virtual', the process of planning what that answer is in capacity terms needs to be ever more rooted in the real considerations of cost, business requirement and service level. |
