The Design And Implementation Of Monitor And Management System For Blade Server

Cloud computing is a new type of information resource management and computing service model, is a new revolution for information industry after Mainframe, PC and Internet.Driven with the data center cloud computing transformation process, how to maximize the utilization of existing data center space, improve performance per watt, and reduce the total cost of ownership(TCO) have become the basic demands of customers. Blade server which have the characteristics of high-density, low power consumption, easy to manage and cost-effective will become an important part of cloud computing data center construction. In order to do monitoring, management, configuration and diagnostics of blade server effectively, need to design and implement the blade server monitor and management system.In this paper, based on the blade server specification, refining and decomposition the blade server monitor and management system design requirements which are the design and development goals. On this basis, the hierarchical monitor and management topology based on two level management controllers (SMC/BMC) was designed. The division of SMC/BMC work is clear, BMC is responsible for blade level management tasks, once blade have problems, BMC will report to SMC actively; SMC is not directly involved in blade level management tasks, but are responsible for BMC scheduling and gathering feedback information, to achieve the purpose of monitor, management and fault detection/recovery of the whole system. The paper also discusses the blade server monitor and management system architecture and working principle from four aspects such as physical architecture design, logical architecture design, runtime architecture design and development architecture design.Finally, the blade power on flow and the key components mechanism are discussed in detail based on "blade power on" typical function from system implementation perspective.