Vijay Swami is a data center technology manager for Presidio Networked Solutions. He has been working in virtualization and data center technologies for more than a decade on both the customer and consulting side.
Presidio Networked Solutions
For enterprises, virtualization means sharing a global pool of infrastructure resources across every application, departmental project and personnel – ultimately leading to significant CAPEX and OPEX savings. However to achieve this, an enterprise needs to look at each individual piece of the enterprise IT infrastructure to determine the best strategy for virtualizing the IT environment. This column will look at the four critical areas where IT directors can achieve the quickest savings from their virtualization initiatives.
• Minimizing Maintenance Time with Desktop Virtualization – Physical desktops are often tough to manage because of the maintenance issues and upgrades associated with the hardware or software running on them. User activity also contributes to wear and tear on parts, and increases the risk of viruses, attacks and operating system configuration issues, costing the enterprise time and money to maintain user productivity. Desktop virtualization helps solve these issues. By using a thin client as the user device and running the actual user desktop as a virtual machine in the data center, IT directors have a single point of entry for OS image patches, maintenance updates, and application deployment. This ensures users have secure access to applications whenever they need them and it streamlines the maintenance process for IT. Also, thin clients aren’t prone to failure, but in the rare event failure occurs, they can be repaired and replaced in minutes vs. hours, increasing productivity throughout the enterprise. Desktop virtualization minimizes the time desktop support groups spend “keeping the lights on” and maximizes the value-added time they can contribute to the business by innovating in desktop technology.
• Greening the Data Center Footprint with Server Virtualization – Server virtualization allows for the consolidation of multiple physical services onto fewer hosts, reducing the space, power, cooling and ultimately, the capital spent on future server purchases – all while minimizing the risk of server failure. Enterprises can easily experience a 20:1 consolidation ratio with server virtualization. However, in the rare event physical server failure does happen, leading virtualization service providers can automatically restart virtual machines by accessing them through another server in the cluster. Furthermore, these virtualized server environments recover in minutes – as opposed to hours or days – offering significant capital savings. Also, by minimizing the potential for long, unplanned application outages, enterprises can save thousands per minute of lost downtime.
• Improving Capacity Management with Storage Virtualization – With traditional methods of managing Storage Area Networks (SANs), most projects never use even half of the space requested. And, by assigning the full amount of requested storage to each project, space is fully used on the array and can never be reclaimed. This wasted space adds up, and when subsequent storage requests follow, additional storage is purchased to match the requested capacity instead of the utilized capacity. Storage virtualization presents the entire storage array as a pool of storage and abstracts disks and RAID groups from the administrator. With this virtual provisioning, the administrator can assign the full amount of requested space to the project, but space is only consumed as it is actually used by the application, allowing for less initial storage and preventing unnecessary storage purchases by using the existing capacity more effectively.
• Eliminating Silos in the Switch Infrastructure with Network Virtualization – Similar to the above mentioned storage issues, where silos of storage set aside for specific applications prevent appropriate capacity planning, the fibre channel switch infrastructure is separate from IP data switching, creating network management issues. Because switch ports are not consumed symmetrically, there may be a need for more storage switch ports, where only IP data switch ports are available (or vice-versa). Network virtualization, also known as converged networking, introduces Data Center Ethernet (DCE) as well as new protocols such as Fibre Channel over Ethernet (FCoE), allowing one 10 Gigabit Ethernet switch port to pass both fiber channel storage and traditional IP traffic, while simultaneously permitting the segmentation of the traffic to occur via software. With these new infrastructure solutions, enterprises can purchase a single switch (vs. three) to handle the management, production data, and fibre channel traffic. This reduction in switches, combined with the reduction in cabling, offers huge savings in capital and operational expenses.
If an enterprise wishes to gain the most return from its virtualization initiatives, it is important to take this holistic approach to virtualization. The market is filled with products to help execute these four initiatives, and by choosing the best solution for each piece of the storage and network infrastructure puzzle, your enterprise will take the first steps towards successful virtualization.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.