Shulou(Shulou.com)06/01 Report--

Data center monitoring management system design 3.1 Introduction Data center has experienced four functional stages of development and evolution, from the early "data storage center" stage, through the "data processing center" and "data application center" two stages, and now has entered the "data operation service center" stage. Data centers at this stage have become an indispensable part of most enterprises or organizations to meet basic business operations and achieve business strategy. Simply put, no matter what industry you are in, traditional manufacturing or emerging services, your data center is tightly tied to your main business. The operation status and planning development of data center directly affect the business development and business operation of enterprises or organizations. Depending on the logical location of the infrastructure in the data center (Figure 1), the operational status of the infrastructure of the data center is also one of the key aspects reflecting the operational status of the data center. Data center construction specifications at home and abroad require monitoring of data center infrastructure; the purpose is to ensure that the operation status of data center infrastructure meets the normal operation and business continuity of various application systems supported and served by data centers. A breakdown of the data centre as a result of a failure of the data centre infrastructure would result in a standstill of the organization's operations. In recent years, banking, insurance, securities, civil aviation and other industries have appeared a number of data center failures, resulting in great social impact and economic losses. Figure 1: Basic implementation in the data center architecture positioning data center infrastructure monitoring management system is a computer software technology based on the use of network technology, database technology, communication technology, embedded technology, industrial automatic control technology, new sensor technology and other professional, automated, intelligent integrated monitoring management system. The system can collect real-time operation parameters, working status and alarm information of various monitored equipment, monitor intelligent and non-intelligent equipment, accurately realize remote signaling, remote control, remote control and other functions, and ensure the normal operation and rapid recovery of various infrastructure facilities of the data center. In addition, on the basis of facility monitoring, it can provide basic data and process support for IT service management. This paper mainly describes the general data center infrastructure monitoring management system construction goals, design principles, monitoring scope, software and hardware architecture, system composition and functions, as well as monitoring management system key technologies. 3.2 The monitoring of data center infrastructure is a basic and core content of data center operation and maintenance management, because on the one hand, infrastructure provides basic and underlying physical environment resource guarantee for data center operation and maintenance, on the other hand, monitoring data and monitoring information for infrastructure operation will become the input of some processes of data center service management (ITSM)(such as ITIL event management, capacity management, etc.). Therefore, standing at the height of IT services, it is very necessary to design infrastructure monitoring systems. Design objectives are derived from the balance between the strategic needs of users and the current use needs. Different customers have different design goals. Determining design goals is not so much a technical problem as a management decision. But generally speaking, we should first grasp the problems and urgent needs faced by users and establish its minimum construction (design) goals. According to my experience, the following points should be taken as the minimum design objectives of infrastructure monitoring management system: (1) to be able to provide real-time status monitoring of infrastructure in all data center rooms. To ensure that the monitoring coverage meets the requirements of data center management, it is necessary to fully cover online real-time monitoring of all power and environmental systems supporting continuous operation of enterprises or organizations. (2) Minimize the risk of manual monitoring and management introducing missed detection and false positives. It is best to change the manual on-site monitoring operation from normal work to abnormal work after the monitoring system is online, for example: only manual (maintenance personnel or equipment manufacturers) regular (monthly/quarterly) on-site inspection (sampling). At the same time, monitoring events are notified (dispatched) in an automated manner to eliminate manual false positives and delays. (3) Provide a closed-loop management of monitoring-control (control)-re-monitoring Monitoring is not an end, nor is it the end of the monitoring system. When the monitored object is in abnormal state, intervention control (manual or system linkage) is necessary. At the same time, the effectiveness and results of intervention control must also be judged by re-monitoring. This is a closed-loop management process, and the system should satisfy. For example: monitoring of air conditioning systems. (4) Relatively improve the work efficiency of the operation and maintenance team. The human resources of the data center are always "stretched" in most enterprises or organizations. Because he is "historically" not a front-line business department, he does not directly create enterprise value and profits. Therefore, we often face the work situation of "more things and fewer people." By adopting centralized monitoring management, remote monitoring management, unattended, automatic inspection and other technologies, the work efficiency of infrastructure maintenance personnel can be relatively improved. This is the most popular project benefit. (5) Being able to record historical operational data of monitored objects One of the values of historical data is that it provides a reference for new problem solving, just like the patient's past medical records. On the other hand, it can "restore the truth," which is the basic requirement of system security audit. This design goal raises the issue of big data storage, processing, and access. (6) Define and report the operation status indicators and health indicators of the monitored objects and even the entire data center infrastructure. The infrastructure of the data center includes many subsystems and many types of equipment. If the monitoring indicators and health indicators cannot be defined in advance (or can be defined), then the maintenance personnel will be submerged in the ocean of waste data and cannot accurately judge the overall operation status of the infrastructure implementation. During the lifecycle of the data center infrastructure, health metrics for the infrastructure should be defined (e.g., average equipment age, average equipment failure coverage, etc.) to quantify the remaining capacity of the infrastructure. (7) To predict the operation trend of the monitored object as much as possible, and prevent problems from happening. It can be imagined that once a major failure occurs in the data center infrastructure, the business of the enterprise or organization will be affected or even interrupted immediately, and the loss will occur immediately. (In 2011, a domestic insurance company stopped its business due to power supply system problems and lost about 300 million yuan). Therefore, if the monitoring system can deduce the problem trend according to the historical operation of the monitored object, it can take measures to eliminate the hidden dangers in advance. In this sense, the most valuable monitoring systems are those that prevent problems. (8) Provide a plan for infrastructure emergencies. Problems will always happen in reality, even if it takes years! So, buy insurance for your surveillance system and develop a variety of problem-solving plans. (9) The monitoring system itself operates without obstacles. This is not to say, just like once the traffic electronic eye is broken, how do you know who ran the red light? In addition, in recent years, with the vigorous development of the data center industry and the functional transformation of IT services, the number of computer rooms has increased dramatically, the scale has expanded, the structure has become more complex, and the monitoring business has grown. New monitoring and management requirements have emerged. Therefore, the new generation of monitoring and management systems should also take the following points as design goals: (10) Meet the requirements of rapid expansion of data centers (11) Meet the requirements of cross-regional networking monitoring (12) Meet the requirements of hierarchical and regional management (13) Meet the requirements of remote disaster recovery (14) Be able to integrate with other information systems of data centers and asset management systems of data centers, or integrate with CMDB and knowledge management systems under ITIL framework. Integration with problem management in the Service Operations module under the ITIL framework. Integration with unified authentication system in data center. Wait a minute. (15) Ability to scientifically evaluate data center infrastructure Based on monitoring infrastructure, further evaluate its level and continuous availability as a basis for data center infrastructure maintenance. For example: data center energy consumption assessment. Calculate power usage effectiveness (PUE) by monitoring the electricity consumption collected by the system. For example, average cost.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.