The development of science and technology makes more and more extensive application of computer, has entered the network era today, advanced technology will function independently of the computer through the network software effectively link network system increasingly large and stable operation, it depends on whether there is a perfect function of network software. Therefore, only to strengthen the daily maintenance of the network, timely detection of problems, timely treatment of the problem, in order to ensure the stability of the network operation, safe, efficient. So do this work, has become the most important in today's network environment.
Common computer network fault classification and impact
Only to find the right remedy, in order to effectively solve the problem. The fault network appears only to accurately analyze and solve the problem of diagnosis, to ensure the stable operation of the network security, which requires professional personnel with rich knowledge of the network, such as network configuration and operation, able to skillfully use diagnostic tools to conduct a comprehensive, in-depth analysis of the fault phenomenon, find from the root of the problem, provide the basis for the smooth solution, to take timely and effective measures to eliminate the fault and restore the normal operation. There are several general network fault: physical device physical layer connection failure or hardware and circuit itself; configuration of network device interface data link layer; network layer network protocol configuration or operation error; equipment performance or traffic congestion problems on the transport layer; the three layer CISCO or IOS network application error.
Suitable for network fault diagnosis of a variety of tools: can use the diagnostic command router itself, can also use network management tools, can also use other diagnostic tools, diagnostic tools provided by CISCO can be ruled out most of the network fault. When a fault occurs, the beginning of the first to view the routing table, you can save a lot of trouble. Secondly, the use of Cisco debug, show command or ICMP trace, ping command can get useful information to help the fault diagnosis and exclusion.
In the collection of information, we have to choose the right order to facilitate the acquisition of information. For example, the ping command is a common method to determine whether the device can be achieved through the IP protocol. For example, the show interface command to carry out information on each interface to check for collection, troubleshooting, and information to obtain the buffer size of use, as well as the use of the specific status of show can make use of the buffer command, if you need to get the memory the case you need through the show proc MEM command or Show the proc command.
The use of each command should be targeted, select the right command to solve the problem with half the effort. The user should pay attention to the data collection, the event of failure, can provide the basis for diagnosis of the problem above, in the computer network problems, the correct use of diagnostic tools and specific diagnostic methods, will contribute to the maintenance personnel to accurately find the cause of the failure, as soon as possible treatment of fault.
Two, according to the nature of the network fault and processing
1 hardware failure
In the process of computer operation, often encounter the plug is loose, the connection is not stable to stop running, or electromagnetic interference lines caused by network abnormal operation or line was unexplained damage resulting in network failure, the fault is also referred to as physical failure, hardware failure often referred to. For example, in the use of the network traffic will sometimes appear unexplained a line on the fall, or the alarm interface popped up, this is due to line disruption to the management information on this line on the port located in this system paralyzed, unable to use, this is a physical fault the most direct and obvious reflection.
The treatment of this problem: first determine the cause of the failure, you can use the ping command to check whether the line is closely connected with the server port, and then take appropriate measures. No connectivity, see if it is loose plug problem, if it is seated on the line, put tight after still does not solve the fault, then there may be a network management center at the end of the plug is loose, can replace a network cable to solve problems.
There is a connection in the line, the operator due to their lack of network skills, resulting in the wrong connection this physical obstacles appear improper operation, another situation is not due to cable meet two routers connected directly, resulting in wrong connection.
The solution to this problem is to: first of all to ensure that the plug in accordance with the specification, followed by the specification of that familiar with the network plug, distinguish between different colors representing the meaning of the cable, the plug connection is correct, to avoid trouble. Like the above mentioned two routers in order to direct connection, it is necessary to meet the quality of the cable line. Similar connection fault is very hidden, the solution to the problem depends on the rich experience of network administrators.
2 logic failure
Configuration error is a common problem in the logic of the fault, the root cause of this problem is the network device configuration is not up to the network failure or abnormal. There are many aspects: setting the parameters of the router port is not accurate enough, the router lost contact with remote address cannot docking; in the network settings, mask incorrect problems and so on, in the fault diagnosis network, first remove the physical fault for this reason, it is necessary to consider the possible configuration error caused by network fault or abnormal.
It is the key to solve the problem that the traceroute command can be used to diagnose and determine the real cause of the failure. If you find that there are two consecutive IP addresses alternating in a band, then the IP packets will be repeated on this line, you can change the configuration to correct the error can be restored. In order to prevent the recurrence of similar problems, the process of handling the fault will be written into the log, leaving a record.
The system load is too high, it is difficult to bear, and the process of the system is also closed. In the process of network operation, often encounter a router SNMP off, the impact on the data acquisition, the network management system can not pass through the router to collect any data, so that the router lost control. Because if the port closure leads to failure, only need to restart this port can be connected to the recovery line, if the router to bear load, such as the inspection found that the temperature is too high, and the use of memory insufficient margin rate is too high, will not have a direct impact on the network connectivity, but for the network to provide the the quality of service will have a certain impact, especially easy to make computer hardware facilities damaged to varying degrees, resulting in the loss of information and other consequences.
Three, according to the network fault object division and processing
1 line fault
The computer needs to be connected with the line, whether the line is smooth network is a strong guarantee for normal operation. Failure is common in the line. Through the Ping check, the main inspection of the port is not a normal response, you can also detect the flow on this line to see if there is still traffic. After the inspection found no flow, or the port of the router is not responding, the problem is caused by the lack of line. Through the ping command to check if it is found that the proximal port port, the down state, can further check the plug, plug loose tight on it; if not found in the proximal port, is the distal port closing problem, promptly notify the other party to deal with the fault, smoothly done or easily solved the problem.
But in addition to this, also because of the problems caused by the line itself line barrier, such as unexplained cut off, disconnect the line or replace can solve the problem; there is a configuration error, such as routing loop, re correct configuration can make the fault process, the network returned to normal.
2 router failure
In the network operation, most of the failures are inextricably linked with the router, so the router is unified as a router. This is the most important in the line fault. One is only a certain relationship with the router itself, typically using high computer CPU rate due to CPU is too hot, too high temperature will cause the burning of the router, cause more serious consequences, it is also the most dangerous situation. The utilization of CPU is too high and the memory margin is too small, although not too dangerous, but it will have a direct impact on the quality of service provided by the network. Detection of router failure, using the tool is the MIB variable in the browser, network system, professional testing through the important data for the router, found that the problem will give alarm by this special management process is responsible for.
Therefore, after find the root of the problem, solve the fault caused by the router is an easy job to do, as long as the timely update of the router can avoid these problems, in addition, re planning is also a good method for network topology. The router's own configuration is a common fault in the router. There is a common type of agreement is not correct, not because of the problem of port, the less likely, so to solve without too much trouble, as long as the choice of the appropriate router when the initial use of the problem smoothly done or easily solved.
3 host failure
Host troubleshooting in all the fault in trouble, the main problem is that the host configuration is not reasonable, the conflict caused by other host IP address and host, the network is difficult to run normally, it is worth noting that will cause the IP does not exist in the network range, the host will be in a state of paralysis, can not be connected. The second is the server settings obstacles. E-Mail is not normal hair caused by improper E-Mail server settings.
Another is the host security failure, for example, the host does not control its finger, RPC, rlogin and other redundant services. In addition, there are a host of other faults, such as improper sharing of the local hard disk, etc.. It is not easy to find the fault of the host and the other faults. The general approach is to monitor the host traffic, in order to detect anomalies, you can also scan the host port and services, to do preventive measures.
Once the preventive measures are meaningless, the host is attacked, it is necessary to timely analysis, and notify the network administrator. Now, install a firewall to protect network security has become the first choice, but not if the firewall in no danger of anything going wrong, the address of the permissions, improper, will appear the same connection failure to solve this problem is very simple, only need to pay more attention on the line in the firewall settings.
Four, conclusion
Today, science and technology more and more developed, accelerate the upgrading of computer technology, but the failure caused by network security is becoming more and more complicated, the text is just from the daily and common fault is discussed and processing method. For specific network technology, should also be specific analysis, in practice, according to the actual situation to accumulate experience. In addition to the establishment of the network management technology, we should pay more attention to the network security technology, from A to Z do understand protection, processing method, network, network to provide a smooth and safe environment, to better promote the development of the network.