Welcome!

SOA & WOA Authors: Elizabeth White, Imran Akbar, Sharon Barkai, Michael Bushong, Marten Terpstra

Related Topics: Virtualization, Java, SOA & WOA, Cloud Expo

Virtualization: Article

Importance of ‘Proof-of-Concept’ – in Right Sizing the Infrastructure

Importance of undertaking proof of concept (PoC) to examine the viability of an approach

It is quite commonly observed now-a-days as a common practice; most of the companies invest a lot of time in engaging consultants & designers and spending colossal amounts of money in capacity planning to size the infrastructure for their specific needs. Not denying the fact that people, capacity planning tools are always helpful to help identify the required amount of resources to size the infrastructure correctly. However, need to consider the fact and it is absolutely necessary to do "Proof-of-concept" especially while making imperative decisions.

There are always concerns raised in terms of obtaining a satisfactory performance. Moreover, mergers and acquisitions have brought in their share of complexity to the existing environment, resulting in technology -vs- application compatibility related challenges. Nevertheless, this may be applicable for setting up new infrastructure for a business critical application from the scratch or for specific IT requirements (for example - data center consolidation, virtualizing a system or going for cloud based solutions). Proof of concepts helps companies in deciding acceptance criteria, right sizing the infrastructure according to their specific needs. It helps in achieving business objectives by controlling the budget over run and helps IT management to plan for cost and procure resources accordingly to ensure successful completion of a project. As the design phase is responsible for many critical decisions, many cost overrun causes are related to such phase. It is identified that most significant causes of cost overrun related to the design phase are due to blindly following the theoretical evidence or by going with by completely trusting on the metrics obtained using unreliable capacity planning tools.

The purpose of PoC is to showcase the benefits using real world end user scenarios and by calculating the TCO for individual cases. Considering the Key system performance base metrics - Processor, Memory, Disk and Network, usually the work loads are classified in to three types (1) Typical user (2) Power user and (3) Advanced Power user. It is always a good practise to calculate load / system usage based on "Power user". If funds permit, it would be even better to use the upper bound for the calculations by considering "Advanced power user" usage in to the account.

PoC helps in determining and size accordingly based on the Average and Peak loads. It enables the consultants in deciding anticipated future growth and leave sufficient room for all key system performance metrics discussed above.

Gartner predicts that the portion of organizations using cloud services will reach 80% by the end of year 2015. Whilst the Cloud Disaster Recovery Service becoming popular these days, companies want to have quick recovery of vital applications in case of failures, by taking advantage of cloud based DR solutions. Hence, it is becoming imperious for organisations to set their own PoC strategy, choose their own POC clouds, navigate technical hurdles & compatibility related challenges, and measure success.

In conclusion, to successfully execute a project, an organization has to give maximum importance to "Proof-of-concept", which defines its success criteria. The use of a proof-of-concept template can be applied to various projects that can help Businesses bridge the gap between the visionary and delivery stages of production efforts.

Fig Illustrates: Resource equals money

More Stories By Sathyanarayanan Muthukrishnan

Sathyanarayanan Muthukrishnan has worked on and managed a variety of IT projects globally (Canada, Denmark, United Kingdom, India) and interfaces with business leaders in the implementation of systems & enhancements.

  • IT Operations Management.
  • Strategic IT strategic road map planning & execution.
  • Data Center Management.
  • Architecture, Analysis and Planning.
  • Budgeting, Product comparisons: Cost - benefit analysis (Hardware, Software & Applications).
  • Disaster Recovery Planning & Testing.
  • Microsoft Windows & Unix Server farms management.
  • Databases (SQL, Oracle)
  • SAN/NAS storage management - capacity planning.
  • Virtualization & Cloud computing (Certified: Citrix, Vmware, Hyper-V)
  • Networking & IT Security.
  • Process refinement, Issues trend Analysis & solutions, ITIL (Change & Problem management)
  • Best Practices Implementations & Stabilization initiatives.