Converged Infrastructure Topologies

Figure 1 shows the major converged infrastructure topologies. The “circle” can be drawn around different elements of the stack. The more the circle encompasses of the infrastructure stack, the fewer the number of SMEs* that need to be maintained.

Figure 1 Converged Infrastructure Topologies
Source: Wikibon, 2013

Wikibon interviewed a number of Wikibon practitioners about the value propositions they derived from different converged topologies. Wikibon found that there is no “wrong” converged infrastructure topology. There is a space for each type within most data centers. However, the feedback was that the fewer the number of SMEs, the greater that value to business and IT.

Business Value from Converged Infrastructure

For the business, the value of IT comes from deploying applications. Two major values drive converged infrastructure:

1. Speed to deploy an application,
2. Speed to update functionality of an application without impacting operational service level agreements (SLAs).

When the reasons for using converged infrastructure were investigated deeply, the root cause for deployment was a compelling case from the business for rapid implementation. This direct pressure from that business case caused IT to find new and innovative ways to implement new applications. Examples researched by Wikibon include:

• The requirement to expand into new international markets, where international law or compliance oversight requires that data stay within strict geographic boundaries, and where speed to deploy is of paramount importance. The solution is to package the solution and “parachute” in a converged total software stack. The solution has been effective from a business perspective because of the very low time-to-result and from an IT perspective because of the low cost of total systems support.
• The requirement was to reduce the time spent by high-value mobile specialists logging into their system at every short stop. With a VDI solution specialists are operational in seconds rather than minutes when they entered the room. The increase in productivity was so high that every location needs to be converted within six months. The benefit to the business is demonstrable increased user productivity. The benefit to IT is the lower cost-of-maintenance from a server managed solution compared to a distributed client environment.
• A large manufacturing company had grown by acquisition and had a large number of different ERP financial implementations. The business effect was that a large number of financial analysts were required to merge the results every month, quarter, and year-end, and the company was not able to react fast enough in a changing market. The solution was to create a single universal instantiation. The value of this approach will equal 50% in IT costs and many times that figure in reduced financial analyst support. The savings are from cost-only functions just keeping the lights on and not adding increased value to the business.
• Response times for certain critical financial analytic applications were getting in the way of the rapid decision-making required by customers. The normal storage infrastructure group could not solve the problem (a combination of high bandwidth and high random performance). The IT solution is a converged infrastructure solution which integrates all components of the stack and provides a separate integrated support group for this set of applications and infrastructure on the “Dev/Ops” model. The business benefit was a significant increase in business from the users of the financial application.
• The health-care industry requires applications and the infrastructure that supports them to be certified. By using converged infrastructure and packaging up the total solution, a services organization was able to reduce the time required to deliver the solution. IT was able to reduce the costs of delivery and maintenance significantly for itself and its customers. The key was creating a Single Managed Entity for the whole stack, including the application.
• An ISV was striving to improve the productivity of its worldwide, distributed development software engineers. The solution was a world-wide VDI system that avoided using a fat client to compile and test new versions of code. By creating a standard offering tailored to the developers, the ISV was able to improve the productivity of its developers by more than 20% and achieved a 30% reduction in time to deploy new versions of a functional upgrade.

Figure 2 shows the increase in value as the level of converged infrastructure is increased.

Figure 2 Value of Different Converged Infrastructure Topologies
Source: © Wikibon, 2014

Case Study of using VCE Vblock for SAP Converged Infrastructure

One of the common converged infrastructure offerings is Vblock from VCE, using equipment from EMC, VMware, & Cisco. The Vblock is offered in many flavors. One of the offerings is focused on SAP.

The enormous potential savings of the project -- many millions of dollars -- was based on creating a common SAP landscape for all divisions. The SAP/VCE Vblock offering is not a single SME. However, the database & application offering from SAP is highly integrated. VCE employs application advocates who research and advise their clients on the exact infrastructure required across the total stack for new versions of the application.

As a result, the project was initially deployed in weeks, the first site cut over in less than six months, and the total deployment was less than two years. Each site found that the level of operational cost for supporting the testing and deployment of new versions fell dramatically.

Challenges to Implementing Converged infrastructure

The feedback on the most important challenge to implementing a move to a converged infrastructure were two factors:

1. The cost of creating an integrated offering,
2. The difficulty of changing today’s mainly stove-piped infrastructure IT organization to one where fewer people are focused on the major value-providing applications.

Wikibon observed that with converged infrastructure, the skills go from deep knowledge about one of the storage, servers, or networking areas to a more general capability with a deeper understanding of development and business value. This is a true Dev/Ops environment.

Action Item: In principle, senior IT managers and directors should strive for the highest level of converged infrastructure integration. The lifetime savings in operational, testing, and planning costs should be expected to be in the 80% range, and the time to value much shorter.

Footnotes: *SME is a Single Managed Entity, similar in concept to a SKU, a Stock-keeping Unit