Ask business executives what tools they are focused on and eight out of 10 times they'll sayserver optimisation. It is little wonder why. Most servers are grossly underutilised, processing at a fraction of their potential capacity, while still eating up a nice chunk of your IT budget. By optimising those underachieving servers, you can harness that capacity more effectively.
Sounds good so far, but unfortunately, server optimisation is not as simple as putting more applications and storage on each server. In fact, server optimisation is a blanket term addressing four approaches to IT infrastructure consolidation. Although the optimisation approaches get more complex as you move from top to bottom, the financial and operational rewards also increase. The four approaches are:
-- Logical Consolidation: This refers to an IT department that implements a unified management system across its servers. Although logical consolidation does not relocate or consolidate any servers, it does simplify management and eases the IT workload.
-- Centralised Consolidation: This involves moving servers to one or two locations. Rather than maintaining servers at various branch offices, everything is housed at centralised locations — at the company headquarters or datacentre, for example.
-- Physical Consolidation: With physical consolidation, an organisation reduces the total number of servers by merging the workload onto fewer, more powerful servers.
-- Operational Consolidation: Also called application consolidation, operational consolidation is the most complex. Unlike physical consolidation, an operational consolidation runs multiple platforms and diverse applications on a single server (or cluster). This technique uses partitioning and virtualisation to run many virtual servers on a single machine.
Server optimisation is not an all-or-nothing proposition, and opting for logical or centralised consolidation in the short term does not eliminate the opportunity to achive physical or operational consolidation at a later time. In fact, the opposite is true. Here's a simple and scalable framework:
-- Plan: Begin by taking an environmental profile, which includes the number of servers and applications you have, the amount of disk space and your current utilisation rate. Use this data to outline a plan of action and to determine the potential financial and operational savings associated with optimisation.
-- Phase: As with your other IT initiatives, structure the plan in phases, so the implementation affects discrete parts of the infrastructure one at a time. Phasing in your implementation not only minimises the risk to your infrastructure, it may also help you to fund one optimisation project with the savings from a previous project, reducing the upfront cost of upgrading your infrastructure.
-- Implement: Once you have a phased plan of action with desired performance outcomes, focus on configuration and implementation. Unfortunately, every market and business model has an IT infrastructure to support it, so single-brand, pre-fabricated optimisation solutions will probably not have the customisation required to fit your organisation. Work with an IT provider who can deliver a wide range of optimisation solutions from multiple vendors, and do not hesitate to look at multiple options before committing to one.
— Manage and assess: After each phase of implementation, schedule time to manage and assess the new infrastructure to identify any problems and to measure the financial and operational gains. This will help you avoid any major gaps created during implementation and provide valuable metrics for justifying the continued investment in optimisation.
-- Evolve: Although server optimisation technologies are now fairly mature (many in their third or fourth generations), vendors will continue to make strides in both software and hardware. Periodically review your utilisation and capacity figures to assess whether your organisation could benefit from either the new technologies or the next level of server optimisation.
Finally, do not forget that any of the server optimisation approaches will likely change your datacentre environment. New servers often consume more power and require more cooling than older, less efficient servers. As a result, your smaller number of optimised servers may have different environmental and power requirements than your previous set-up. Make sure you have the power and cooling infrastructure to address the changes.
Ultimately, server optimisation is both a technology and a business decision. Although the benefits to IT are substantial, CDW's research shows that 66% of IT managers consolidate to reduce costs. As business leaders are always looking for ways to improve shareholder value, investing your time and effort in an optimisation initiative is a proactive way of highlighting how IT contributes to your organisation's success.