The Dollars And Sense Of Building A Cloud Infrastructure

By Dan Timko, President & CTO, Cirrity

Numerous studies in the media have touted the “savings” of cloud computing. Unfortunately for managed services providers (MSPs) and other resellers, virtually all those savings are seen at the customer level. As customers become more savvy, resellers must invest in powerful, reliable infrastructure, faster network connections, highly skilled technicians, and other attributes of top-quality cloud providers. Building out a true cloud infrastructure can cost millions of dollars, but it is common for small providers to start by hosting customers on a simple virtualized platform. For this reason, we won’t get into the cost of building a cloud hosting environment. A more important question for organizations is, “How much will it cost to build a cloud environment that will meet the needs of my current and future customers?” Making the wrong choices can quickly ruin a firm’s reputation as a quality provider.

Hardware And Software Investment

I have worked with providers that initiated cloud services with only a few servers, hosted as part of their own server farm. They think, “I’ll start small and offer Infrastructure-as-a-Service (IaaS); most of my customers are small and mid-sized businesses (SMBs), and IaaS is an easy service to manage.”

Not only do they end up with an environment that isn’t scalable to meet customer needs (see “scalability,” below), but they don’t invest in data-center grade hardware, don’t ensure sufficient redundancy, and don’t consider the manpower resource they will need to manage it.

Companies that build a cloud architecture should consider projected future needs and double — or even triple — their most aggressive capacity estimates. They should plan to build, or lease space in, a data center running on enterprise-class infrastructure with a high level of physical security. They should also plan for the day when customers will want more services, such as virtual desktops (what we call Desktop-as-a-Service, or DaaS), disaster recovery as a service, and other as-a-Service” solutions.

Personnel Management

Personnel management is perhaps the biggest problem I see with companies hosting their own clouds. Cloud architects and engineers are expensive resources, and unless the cloud framework is extensive, these resources may not have a full-time workload after the environment is up and running.

As a result, management may try to reduce expenses by cross-training support technicians to manage the cloud environment, leaving them unavailable to make it a priority when problems occur. Alternately, a firm may hire a cloud engineer and then try to maximize the resource by having the engineer perform billable technical work in other areas. As the billables increase, management puts pressure on the engineer to continue that path, and the cloud platform degrades as it is neglected.

By not allowing cloud engineers to remain focused on the cloud platform, companies also reduce the amount of time for R&D projects, which means the company isn’t developing new services that will increase cloud business.

Scalability

Unless a cloud infrastructure is architected properly from the outset, scalability may be a challenge. A small-reference architecture simply doesn’t scale to an enterprise-grade environment. A small cloud can be reconfigured and redesigned to be scalable, but the cost is often prohibitive. It is much better to plan for scalability at the outset.

Security

Security and compliance are two of the greatest challenges cloud providers experience. Even the smallest customer expects enterprise-grade security, and companies planning to host their own cloud environments usually need security experts to help them build them in a secure fashion. From there, they must contract appropriate partners for regular penetration testing, assessments, and other needs. A security breach can literally shut a company down, in terms of both lost reputation and financial exposure.

Moving Up And On

I speak with very few resellers who are just now beginning to build their own clouds from scratch. Most recognize that without the economies of scale that come with a larger platform and customer base, it is cost prohibitive to be competitive in this space. Rather, most are already running some cloud functions and are considering whether they should upgrade their in-house clouds, contract for cloud resources with data centers, or license ready-made, cloud-based solutions such as IaaS, DaaS and DRaaS from a provider that targets the reseller market. Unfortunately, MSPs are losing customers every day to cloud providers with rich portfolios. Furthermore, firms that botch a cloud migration or enable a security breach due to inexperience or inadequate preparation will suffer along with their customers.

Even the smallest customer has big expectations when it comes to a cloud environment. Achieving both service excellence and profitability in the cloud involves careful evaluation and targeted decision making. It requires firms to find a workable model for both themselves and their customers. Nevertheless, resellers that address their limitations, work with trustworthy cloud partners, if appropriate, and successfully make the transition to full-service cloud providers can ride this technology wave into a very bright future.

Dan Timko is president and CTO of Cirrity, a channel-only, Cisco-powered secure cloud service provider with locations throughout the continental United States. To query Timko about this solution for MSPs and other channel partners, shoot him an email at DTimko@cirrity.com. For more information on Cirrity, visit www.cirrity.com.