Would you believe that a large number of desktop virtualization projects get to roughly about 30% completion and then never reach production? An even greater number fall apart prior to even reaching a small scale pilot stage.
What makes ITAMG an expert to even share a list like this? We are often called in to liquidate the results of failed desktop virtualization projects. Our credentials are backed by our warehouses that are full of decommissioned servers, networking equipment and lightly used and new in box thin clients.
With the greening of IT, telecommuting, cloud adoption, securing data at rest, disaster recovery, big data analytics and data center consolidation being all the rage today, here we will list out the top 10 issues that can make your virtualization project fail even before it has begun or hopefully, lead to better planning.
IT infrastructure is known to generate a lot of heat and white noise. Installation of additional cooling systems or upgrading existing circuits and duct work are often overlooked as part of overall project launch costs. Additionally, the amount of money spent on running these systems may outstrip any energy savings if that was a goal of your virtualization project.
Many virtualization projects require both software and hardware from multiple vendors. Ensuring that all components that will be used in your project arrive in time and stay within the overall project budget can frustrate even the most experienced IT professionals. Additionally, the financial counterparts within your organization who need to sign off on new purchases or changes will be better advised for internal needs like sending out payments on time to vendors.
Perhaps the number one problem that halts many virtualization projects are infrastructure costs. After reviewing the return on investment in areas like hardware, set up and maintained, these figures often come in at roughly 70% to 100% higher than the cost of software licenses needed as part of the overall project.
In the early stages of project scoping, questions like will the system run fast enough to satisfy end user demand often come up. Components like servers, storage and network components all needing to run efficiently in order to prevent outages and end user dissatisfaction. Right sizing the system will increase performance and end user adoption.
Excessive power requirement can delay virtualization projects for up to a year while waiting for zoning or construction permits and the actual installations of new circuits or power management systems. Moreover, increasing the overall power consumption attributed to IT operations ironically creates a case for not virtualizing certain systems based on power requirements alone.
The goal of many virtualized systems is to run more efficient operations in order to serve many departments from a common platform. Conflicts may arise when it comes time to allocate costs to different departments and divide workload between dedicated IT teams that often only focus on specific areas like servers, networking or storage technology. In the end, too many cooks in the kitchen may prevent your virtualization project from running smoothly unless it is effectively managed,
Organizations typically decide to either go all in with adopting the latest technology or trial it through small scale pilots through the use of on-hand surplus equipment. Not understanding how a virtualization project may scale to the organization and its end users may be its undoing.
Just because a 42U rack can hold that many servers or network devices, does not mean it will because of cooling, power and cabling requirements. It is beneficial to pay close attention to future space requirements if expansion may be required. Conversely, if too much space is reserved for future IT projects, it may then cut into other business space requirements for an additional conference room, office, cubicle space or employee nap pods.
Estimating ROI on a virtualization project takes a lot of guesswork. Even after laying out your system requirements into one of those fancy vendor configuration spread sheets a few things could happen. After delivery and configuration, you may have over bought, under prepared or just find out that its best to just cut losses on the project before it further spirals out of control.
Consider the load bearing capability of the floor, even if your data center is located in a sub-basement versus on the 32nd floor of a Manhattan highrise. Components are added one by one and adding up the overall weight of any number of 1U or 2U devices may surprise you over time.
IT Asset Management Group hopes that you found this list helpful and we welcome your feedback. If we missed something that you found harmful or helpful in your planning of your virtualization project or decommissioning project, please let us know in the comments section. It will also help other readers benefit too.