The quarterly publication of the International Legal Technology Association
Issue link: https://epubs.iltanet.org/i/27607
I magine that your firm’s top rainmaker calls you into her office and insists that she needs special software in order to entertain new mobile communications business from her most important client. You quickly ascertain that this must-have application will require a substantial backend, including a couple of dedicated servers, a database and lots of storage. Your current environment doesn’t have the capacity for these new systems, and several weeks will be needed to test the software and put it into production. In the meantime, a competing law firm managed to get the same system up and running, and they snagged your firm’s client. The big question is how did the competing firm manage to launch the application before you? They leveraged their private Cloud offerings, which are built upon server virtualization technology from VMware. The new services were provisioned quickly and delivered to the client through the Internet. Cloud Computing and How Does VMware Virtualization Work? VMware virtualization begins with software called a “hypervisor” being installed onto a physical server. It establishes a base environment that allows multiple application servers to be loaded onto the physical server. More importantly, it permits those multiple servers to run concurrently; thus, sharing the resources of the physical server. Since these servers do not directly interact with and control the physical hardware, they are considered virtualized. A virtual server provides the same services as the more traditional approach so the end-user sees no difference. From the system administrator’s point of view, a virtualized server comes with many operational benefits, and for the business- minded, there are cost savings since fewer physical servers need to be purchased and maintained. The process of Virtualization Cloud computing holds the promise of providing application services in a more flexible and timely manner, and without much hassle. Clouds can be public — where the services are available to anyone over the Internet — or they can be private — where the Cloud is restricted to certain users and typically delivered in a more secure manner. In either case, the behind- the-scenes establishment of Cloud services requires a highly adaptable, responsive infrastructure. Server virtualization fits the bill and stands at the heart of providing Cloud-based application services. VMware server virtualization, in particular, has become a key building block in establishing Cloud services. Applications run on hardware-independent servers called “virtual servers,” which exist within a set of physical servers that provide the memory, processing and storage resources for the virtual servers. In short, a physical server may host several guest servers, providing diverse services such as e-mail, database and litigation support. Since the application server is liberated from direct dependency upon physical hardware, there is no longer the one-to-one relationship between the application running on a server and the server box. Instead, using VMware virtualization technology, the physical server becomes a pool of resources available for any number of application servers. “Using VMware virtualization technology, the physical server becomes a pool of resources available for any number of application servers.” establishing a virtualization infrastructure entails first specifying the application servers that are required for the production systems. These specifications account for the number of processors, memory sizing and disk space — just as you would currently specify for a physical server. With these specifications in hand, the next step is to size the virtualization environment. Sizing combines all the parameters to arrive at a number of physical servers, along with their own specifications that accommodate the application servers, while aiming for acceptable performance goals. These physical servers become the host systems, each running more than one virtualized application server; therefore, the hosts should have similar hardware specifications, be powerful and contain ample memory. It is especially important to correctly specify disk space since space must be allocated not only for the application data, but also for the virtual servers themselves. Disk performance must also be taken into account because the speed of disk access will impact performance while running application servers concurrently. Each virtual server competes for the underlying physical disk, so this disk must be sized appropriately to be responsive to the needs of all of the virtual servers. Under-specifying the disk performance will lead to application slowness, which is especially hazardous for disk- intensive applications such as e-mail and database systems. Peer to Peer the quarterly magazine of ILTA 63