It is well established that virtualization can help you take a significant bite out of your IT hardware budget. Ideally, however, that should merely whet your appetite for a bounty of potential benefits offered by a virtualized infrastructure.
Virtualization enables a single piece of hardware to run multiple operating system images at once, thus giving organizations the ability to repurpose and consolidate existing servers. But cutting hardware costs is merely an appetizer. The main course is the business flexibility and agility enabled by virtualization.
With the right tools, virtualization offers a number of other benefits — including dynamic provisioning, more efficient software development, comprehensive staging and testing, real-time migration and affordable disaster recovery. By creating a virtual pool of processing and storage resources that can be quickly allocated to new applications and more easily moved from primary to backup locations, virtualization can make an IT organization more agile, resilient and responsive.
“Adopting an IT philosophy based on virtualization only begins with the adoption of a virtual platform such as VMware,” said FusionStorm CTO Vince Conroy. “You also need a provisioning solution, an automated management solution and a monitoring system to really optimize your virtualized IT infrastructure.”
Fine Tuning
Virtualization has been compared to carpooling: instead of four people taking four cars to the airport, they all share one vehicle. In essence, virtualization management tools help organizations understand which members of the carpool have the most luggage and the tightest schedules.
“It’s one thing to be operating in a virtual world in which servers can share applications and operating systems,” said Marc Franz, FusionStorm VMware Director. “But if I can’t manage those boxes well enough to respond to changes in capacity and utilization, then I’m not taking full advantage of virtualization.”
That is roughly the state of virtualization in most organizations today. Organizations are chiefly focused on high-level benefits such as server consolidation and cutting costs. Going forward, however, organizations must look to fine-tune their systems to ensure that resources are aligned to the needs and priorities of the business.
FusionStorm can perform an assessment of an organization’s virtualized environment and demonstrate how to use a variety of different tools — either those native to the virtualization platform or tools from third-party vendors — to manage systems through a browser interface. With such tools, administrators can manage the state of each virtual server, ensure servers are operational and view a summary of processor and memory resource utilizations. The interface also supports the creation of new virtual systems, provides greater details pertaining to individual system configurations and supports other essential management tasks.
Automated Provisioning
One such task is provisioning — the setup, installation and configuration of servers. This has traditionally been a tedious and time-consuming manual process that opens the door for a variety of configuration errors. Large IT groups, in particular, often wind up with server configurations that vary from group to group and location to location. Software updates and security patch installation can occur randomly throughout the network, which often leads to hardware and software conflicts.
In a virtualized environment, however, organizations can automate server provisioning. Once a virtual server has been created and properly configured, it can be saved as a so-called “gold image” and then deployed multiple times. There is no need to purchase and configure new hardware — the gold image is easily updated and virtual servers can be swapped in and out fluidly based on business needs.
“In a virtualized world, you don’t have to worry about the vagaries of the specific hardware that underlies the operating system, which has always been the bane of provisioning,” said Franz. “In a virtualized server environment, particularly with VMware, every network card, video driver and SAN interface looks the same, every SCSI device looks like a generic device, and every storage array looks the same from the perspective of the storage drivers and connectivity paths. From a configuration perspective there is absolutely no difference.”
Once the final configuration is set, it can be easily modified if a change is required due to spikes in utilization, changes in application functionality, virus contamination or some other external event.
“Once I’ve determined a change needs to be made, it is important to make that change across all of my servers performing that particular function,” said Franz. “So I just need to go back to my gold image, modify it and bring it up on a test virtual server — which costs me nothing to have. Once I’m sure that I have the correct configuration, I can roll that back out across all my servers in a matter of minutes.”
Because servers are provisioned, deployed and configured consistently, there is an extremely low error rate. This also reduces hardware and software conflicts on the network. What’s more, reverting to previous configurations involves only a mouse click rather than a complete rebuild of a physical server.
Automated provisioning can also reduce the cost and complexity of software development and test systems by eliminating the need to purchase test-bed hardware for each discrete development or test environment. Developers can create multiple environments with different operating systems on the same server.
Streamlining Recovery
Disaster recovery is also enhanced by rapid provisioning capabilities. Redundant servers can be easily created by provisioning virtual machines rather than purchasing additional hardware. Recovery of the gold image is a single-step process that greatly improves the ability to meet time-to-recovery targets. And because virtualized servers do the work of more physical servers, organizations can reduce their hardware investment in a separate disaster recovery site.
“Hardware cost has always been one of the biggest challenges with disaster recovery,” said Franz. “If you have 50 servers in a production facility, do you want to keep 50 servers available at a disaster recovery facility at various states of readiness? That is a tremendous amount of overhead.
“But in a virtualized environment, you don’t need as many physical servers in the disaster recovery site. You might only need 40 percent. And you can keep those pre-provisioned — each time you make a change to your gold image, you just push that out to the DR site as well. In the event of a disaster, you just click a button to bring the imaged servers at the DR site online and then you shift the traffic loads over. It not only allows you to reduce the quantity of servers, but to really manage the process of keeping the DR site and the primary site in sync — not just from a data perspective but also from the operating environment.”
Most industry experts say the key to realizing the full potential of virtualization is to establish specific goals and take a phased approach. That’s where FusionStorm can help. After completing an assessment, FusionStorm will provide a detailed document describing where the technology offers the greatest benefits, how to implement the proper management, monitoring and provisioning tools, and how to adjust management styles to fit a virtualized environment.
“The key is to take it one step at a time. We can develop a roadmap to measure where you are currently, what the next steps are and how to reach the desired end result,” said Conroy. “That’s FusionStorm’s strength. We provide the resources, the discipline and the methodology to help you every step of the way.”