銆?1CTO.com Special Report銆慛ot long ago, while interviewing the director of the IT center at an energy group’s Guangzhou branch, I heard this remark: “The ITSM system helps us establish a solid set of IT operations management solutions and methods, but when it comes to actual network issues, manual intervention is still required. This is especially true for a network as large and structurally complex as ours 鈥?troubleshooting and proactive fault handling remain very challenging problems.”
From this IT director’s words, we can see that ITSM (IT Service Management) and NSM (Network Systems Management) have not yet been well integrated. In fact, currently in China, ITSM systems emphasize coordinating the relationship between management personnel and IT users. Simply put, it deals with relationships between people. NSM systems, on the other hand, handle the relationship between network operations staff and network hardware infrastructure, meaning it is primarily responsible for the relationship between people and machines. For an enterprise, both are critically important. To excel at IT operations, a balanced development of both must be achieved.
At the end of 2007, Ctrip and SiteView Technology reached a cooperation agreement to build an IT operations management system. After more than a year of testing, small-scale trials, full deployment, and pilot operation, the Ctrip IT operations management platform was officially accepted by March 2009. The successful implementation of Ctrip’s IT operations management system is a highly influential and in-depth project in the IT operations field and is also one of the largest network operations projects currently in China. Today, Ctrip’s operations team of over a hundred people uses the SiteView operations platform daily to ensure the normal operation of its network and business systems. With the help of this system, Ctrip’s operations team can provide more efficient support services for IT applications handling up to 150,000 calls per day and massive volumes of website queries.
Since the vast majority of Ctrip’s business is network-based, their requirements for network reliability and availability are extremely high.
This is a very typical NSM system implementation case, and what sets it apart is the scale of the project.
It is not difficult for one person to manage a garden, but it is far less easy for one person to manage an entire forest well. By the same logic, if an NSM system manages a relatively simple and small-scale network, it might not be a challenge. However, if it needs to manage a large-scale, complex network, it is absolutely a test for the NSM system.
Many readers may not understand this statement: “Isn’t a network operations system meant for managing networks? Why would a large-scale network pose a challenge for a network operations system?” The answer is actually quite simple. A network operations system is itself a system, with its own front-end pages and back-end database. When the managed network is relatively large and complex, the backend data volume of the network operations system will be substantial, and the number of concurrent polling rounds for each network node will also be high. In other words, managing large networks demands extremely high performance from the NSM system. One of the core technical difficulties of NSM is system performance improvement.
Term Definition: Concurrent Polling Count: Refers to the number of network nodes that the network operations system simultaneously sends commands to for access within the same time period, and from which it receives returned information at the same time.
As for ITSM, the core issue lies in how the system-defined processes adapt to the user’s continuous changes, enabling the ITSM system to flexibly adjust according to user modifications. That is to say, one of the technical difficulties of ITSM is its own flexibility and the rationality of its processes (of course, ITSM systems have many other challenges, which we will not elaborate on here due to space constraints).
Before implementing the SiteView system, Ctrip used a combination of open-source network management software and an ITSM