DGM

Virtual Resource Pool (VRP) is a logical container that is used to reserve virtual resources and to provide guaranteed total computing resources for all virtual servers in clients’ Virtual IT Infrastructure. These resources are dynamically distributed amongst the virtual servers that are running in the pool, according to their current needs. The global parameters of the VRP determine the actual capacity of the clients’ virtual environment.

Each VRP is defined by establishing the following parameters:

• Total number of Virtual CPUs (pcs.)
• Total CPU Speed (MHz)
• Total amount of RAM (MB)
• Total amount of On-line Data Storage Space (GB)
• Total number of Virtual LAN Ports (pcs.)
• Total amount of Backup Data Storage Space (GB)

Virtual servers and extra resources can be added to or removed from the Virtual Resource Pool as needed. When adding new virtual servers to the pool or when increasing the parameters of individual virtual servers, a recalculation of the free pool resources has to be made.

The high-end virtualization technologies that we use allow us to configure a VRP with lower resources than the “mechanical sum” of the individual resources allocated for each individual virtual server, without causing bottlenecks. Since different virtual servers run different workloads and since peak loads are short and almost never happen at the same time, the average VRP utilization is always lower.

Since the monthly fees are based on the parameters of the resource pool and not on the mechanical sum of virtual machine resources, cost optimization can be achieved without losses in performance or quality.

In order to determine the optimal parameters for a client’s Virtual Resource Pool an initial configuration is deployed, based on the initial requirements. This initial configuration then undergoes testing that simulates the actual workload of the client’s environment.

The elasticity of the Virtual IT infrastructure service makes it possible to quickly and easily change the parameters (CPU, RAM, Disk space, bandwidth, etc.) according to the actual load and thus the final parameters are optimized for the actual resource usage.

It makes sense to provide Virtual Resource Pool with lower parameters and later adjust them as needed. This is contrary to the physical world, where one needs to overbuy and oversubscribe resources at first.