IT department of the enterprise may need to operate hundreds of enterprise applications on hundreds of servers. Under many circumstances, every application or case needs a special server, even if the application only occupies a little part of the available physical CPU or memory of the server. The result is that the overall computing resource cannot be fully utilized. Besides, IT department is facing more and more problems such as purchase, deployment, monitoring and maintenance of server and related resources (e.g. power, cooling, bracket space, wiring and supporter).
In order to overcome such challenges of the data center, Rich Cloud Technology provides the server virtualization and related technologies to support the server network and data center administration. Different from the traditional design of the operation of a single application on a single physical server (i.e. 1:1), now a physical server may include many virtual servers (also called virtual machine) and supports various applications by a single device (i.e. N:1).
In order to reduce the operating cost of the data center, Rich Cloud Technology systematically considers several elements in server virtualization deployment:
Total cost
An enterprise shall firstly calculate the I/O cost to include the cost of server, exchanger, network card and cable, as well as power, bracket space and time for infrastructure administration that the enterprise may discover the neglected cost saving space. Seriousness of the problem must be firstly understood before understanding the cost saving space.
Estimation flexibility
Does enterprise demand regularly change? Does the enterprise consider administrating the IT system operation? If so, it will be profitable to elevate the flexibility. When distributing a new linkage to the server, the enterprise may obtain more profit. When estimating the interconnectivity demand of the device, unexpected expense must be put into consideration. If any server may obtain any resource at a time point, the enterprise may more rapidly respond to the unexpected change to better meet the demand.
Bottleneck discovery
Is server I/O of the enterprise limited? It may reduce the device utilization rate and cause unnecessary cost. If the enterprise possesses any monitoring tool such VMwareESX device, I/O utilization ratio shall be checked that the enterprise may find it as a limiting factor, thus further deepening the integration.
Simplification
Since I/O cost is quite close to the server cost, it shall be simplified in the next step of configuration. Purchase and installation cost of the simplified I/O infrastructure will be reduced and also the risk of shutdown caused by artificial mistake will also be reduced.
Cost saving method discovery
Cost saving of one point may bring about the possibility of cost saving of another point. Simplified I/O infrastructure will make the server smaller and cost less to save more expense for each server.
Innovation
The same method in the future means that cost structure of the enterprise will not change in the future that the enterprise has to find out the high-cost field that can be innovated. Technical innovation (e.g. server virtualization, iSCSI storage and I/O virtualization) provides more cost saving methods than the compulsory saving. Core of the virtualization is: cost the least to do the most.
Trial running
Investment on the new technology needs time to be returned while the cost saving becomes effective rapidly. In order to accelerate this process, the new cost saving method shall be carried out in some IT environments of the enterprise for the enterprise to optimize such method in the large-scale popularization. Besides, daily data of the enterprise proves that such method is quite beneficial, and the enterprise itself may obtain the experience in guaranteeing the success.
Case Study：Data Center Platform Construction for a Large Manufacturing Enterprise