What is centralized monitoring

With the continuous advancement of the global informatization, the division of labor in the IT service industry has become more and more elaborate and clear. As the basis of all IT services, the data center and related infrastructure are directly related to the normal, continuous and stable operation of the IT service system. Any decrease in efficiency or failure of any part will result in a decrease in the availability of IT services, resulting in poor access to information, and in a variety of unpredictable and significant losses.

Centralized monitoring and promotion

How to increase the availability of data centers has become one of the important topics in the "high availability IT services". As the first line of defense to ensure the availability of data centers - "Centralized monitoring" can quickly help companies achieve "high availability" goals.

The significance of centralized monitoring

According to the definition of ITIL, the so-called "availability" refers to: "The ability of a configuration item or IT service to perform the agreed functions as needed. Availability depends on reliability, maintainability, serviceability, performance, and security. Usability is usually Calculated as a percentage. This calculation is usually based on agreed service time and downtime." High availability IT management refers to improving the availability of IT systems through the improvement and optimization of key elements of high availability such as IT architecture and operation and maintenance management, infrastructure and management, disaster recovery and construction, operation and maintenance, and security and management. To better protect the business continuity and innovation process.

In terms of measuring availability, it is specifically divided into three different indicators such as MTTR/MTBF/MTBSI. Whether it is MTBSI or MTTR, there is an important component - "Detect time (detection time)." It can be seen that whether or not the failure of each management object in the data center can be discovered in a timely and effective manner through effective monitoring and management has formed a sufficient condition for high availability of the data center.

The role of centralized monitoring

The management objects of the data center mainly include two parts: infrastructure and IT infrastructure. Among them, infrastructure includes power supply and distribution, UPS, air conditioning, fire protection, security, environmental monitoring and other computer room systems; the infrastructure includes network equipment, host equipment, storage equipment and other IT equipment.

The goal of centralized monitoring is to be able to monitor the operation of the infrastructure and IT infrastructure through the application of management and technology, real-time discovery and notification of failures and anomalies, and also to collect and organize monitoring data for capacity. Management, incident management, problem management, and compliance management provide the basis for analysis and ultimately achieve the goal of high availability in the data center.

Centralized monitoring management

With the development of technology, many third-party monitoring tools have begun to appear. These tools can realize centralized data collection across devices, cross-platform, and cross-system, and can also set corresponding thresholds for different monitoring objects. Finally, they can also be realized. Unified display and alert. With the advent of these tools, IT managers can discover the failures of managed components in a faster and more accurate way. As a result, valuable time was saved for the repair of the fault and the restoration of the service, which increased the availability of the entire infrastructure.

Capacity management

The monitoring management will also use the performance collection function of the monitoring tool to monitor the key performance points of some key applications and obtain the performance data of these key points to evaluate the capacity of the IT system. When there is a deviation in the capacity plan of the IT component's performance, the performance of these organizations can be expanded in time to reduce the possibility of business interruption due to insufficient performance.

Safety management

Monitoring and management can use some security monitoring tools to check the security situation of the components and the compliance with the compliance requirements during operation. Some GDS partners, for example, use some security software to perform real-time log collection and security analysis on firewalls, antivirus and intrusion detection devices, and compare their security policies or security standards to help data center managers in data centers. The rapid positioning and problem analysis of security issues in operation.

Monitoring tools

The purpose of monitoring and management is not to monitor the tools themselves, but to find infrastructure and infrastructure problems in a timely manner through manual or technical means. In accordance with the established requirements, the identified problems are mobilized according to established management processes and tools. The involvement of technology and management personnel will ultimately solve problems that may occur in the data center, such as events, capacity, and availability. Therefore, how to make data center staff aware of problems reported in the monitoring tool, how to implement the follow-up management process, to avoid misstatements, omissions, become an important challenge for monitoring and management.

Business monitoring

The data center provides information services. It can also be called business services. Monitoring an individual device independently can no longer meet future needs. For managers, what concerns more is whether a service provided by the data center can run normally. Therefore, future monitoring solutions need to start from the business and service level and separate physical devices. , closely associated with the business to form a business device view, the availability of each device can reflect the availability of the business.

Virtualization Cloud Monitoring

Virtualization is the trend of data centers in the future, but it is difficult for monitoring tools to distinguish whether the monitored server is a physical machine or a virtual machine. It is also impossible to know that the hardware system will have a potential impact on server availability, and the virtualization platform Availability directly affects the availability of virtual servers running on it. The monitoring software should deal with the problems of the main server hardware. However, if the main server is in danger, any virtual machine running on the host faces the same problem. Therefore, the high availability solution for the virtualized cloud environment will also be One of the future trends.

Impact Analysis Model

The basis of business monitoring and virtualization environment is that different devices can establish clear management and form a network of devices and devices. This requires the establishment of a CMDB (Configuration Management Database), which clearly describes the information attributes of each device. , and the relationship between devices. By establishing a CMDB repository, a business impact model is formed. For example, the following is an impact modeling of an online trading system:

In the trading system's impact model, for example, if the "storage" fails, it directly reflects that the service is unavailable, and that "online trading system", "database server", and "online trading system" are all unavailable, according to Dependent on the analysis, you can directly locate the cause of the failure, thus avoiding system-by-system failure analysis.