Network failure; Common faults; Classified diagnosis; Physical failure; Logical failure i. Physical failure
Physical faults generally refer to physical problems or hardware problems of lines or equipment.
(1) line fault
In daily network maintenance, the incidence of line faults is quite high, accounting for about 70% of the faults. Line faults usually include line damage and serious electromagnetic interference.
Troubleshooting method: If it is in a short distance, the simple way to judge whether the network cable is good or not is to plug one end of the network cable into the RJ45 socket of the host computer that can normally connect to the LAN, and the other end into the normal HUB port, and then Ping the host computer or router at the other end of the line from one end of the host computer, and judge according to the on/off. If the line is a little long, or the network cable is inconvenient to adjust, use the network cable tester to measure the quality of the network cable. If the line is very long, such as provided by suppliers such as post and telecommunications departments, it is necessary to inform the line supplier to check the line to see if it is disconnected halfway.
For the investigation of whether there is serious electromagnetic interference, we can use the shielded wire with strong shielding to test the communication of this section of the network. If the communication is normal, it means that there is electromagnetic interference, and pay attention to stay away from objects with strong electromagnetic fields such as high-voltage lines. If it is abnormal, the line fault should be eliminated and other reasons should be considered.
(2) Port failure
Port failures usually include loose plug and physical failure of the port itself.
Troubleshooting: This kind of fault usually affects the signal lights of other devices directly connected to it. Because the signal lamp is intuitive, the occurrence range and possible causes of the fault can be roughly judged by the state of the signal lamp. You can also try using another port to see if the connection is normal.
(3) Hub or router failure
Hub or router failure here refers to physical damage, which can not work, leading to network failure.
Troubleshooting: Usually, the simplest method is to change the troubleshooting method, and connect the hub (or router) with the network cable and the host computer, so that the communication is normal. If the communication can be normal, the hub or router is normal; Otherwise, switch the hub port and check whether it is a port failure or a hub (or router) failure; Many times, the indicator light of the hub (or router) can also indicate whether there is a fault. Under normal circumstances, the light of the corresponding port should be green. If communication fails all the time, it can be considered that the hub or router is faulty.
(4) Physical failure of the host computer
Network card failure, the author also classifies it as a physical failure of the host, because most network cards are installed in the host, and the configuration and communication are completed by the host, which can be regarded as a network terminal. This kind of fault usually includes network card looseness, network card physical fault, host network card slot fault and host itself fault.
Troubleshooting: I won't go into details about the host's own faults here, but only introduce the situation that the host and the network card can't work together. The best solution for host network card looseness and network card slot failure is to replace the network card slot. In the case of physical failure of the network card, if the above replacement slot can't solve the problem all the time, take it to another normal host to test the network card. If it still doesn't work, it can be determined that the network card is physically damaged, and the network card can be replaced.
Second, logical errors.
The most common situation of logical failure is configuration error, that is, network abnormality or failure caused by network device configuration error.
(A) the router logic failure
Router logic failures usually include incorrect setting of router port parameters, incorrect router routing configuration, high utilization rate of router CPU and low memory margin of router.
Troubleshooting: The router port parameters are set incorrectly, which will cause the remote address not to be found. Use Ping command or Traceroute command (route tracker: in UNIX system, we call it traceroute; Tracert in MS Windows), check which node of the remote address has problems, and check and repair the parameters of the node.
Router routing configuration is wrong, which will make the routing cycle or the remote address not found. For example, two routers are directly connected, the exit of one router should be connected to the entrance of the other router, and the entrance of this router should be connected to the exit of the other router. At this time, the network cable made must meet this characteristic, otherwise it will also lead to network errors. Traceroute tool can be used for this fault. After finding a certain segment in the Traceroute result, two IP addresses appear circularly. At this time, it is usually the far end of the line that points the port route to the near end of the line, resulting in repeated delivery of IP packets on the line. The solution to the routing cycle is to reconfigure the static route or dynamic route of the router port and set the route to the correct configuration, so that the line can be restored.
The router's CPU utilization rate is high, and the router's memory margin is low, which leads to poor network service quality. For example, the smaller the router memory, the higher the packet loss rate. It is intuitive to detect this kind of fault with MIB variable browser, which collects router routing table, port traffic data, billing data, temperature and load of router CPU, and remaining memory of router. Usually, the network management system has a special management process to continuously detect the key data of the router and give an alarm in time. The only way to solve this problem is to upgrade the router, expand memory, or re-plan the network topology.
(2) Close some important processes or ports.
Due to the influence of system or virus, some important processes or ports related to network connection data parameters were unexpectedly closed. For example, when the SNMP process of the router shuts down unexpectedly, the network management system will not be able to collect any data from the router, so the network management system loses control of the router. Or the line is broken and there is no traffic.
Troubleshooting: Ping the port near the line to see if it can be pinged. If you can't ping, please check whether the port is closed. If so, the port has been closed, resulting in failure. At this time, just restart the port and the line can be restored. 3) host logic failure
The network failure rate caused by host logic failure is high, which usually includes improper installation of network card driver, conflict of network card devices, improper setting of host network address parameters, improper installation of host network protocol or service, and host security failure.
1. The network card driver is not installed correctly. Improper installation of network card drivers, including not installing network card drivers or incompatible wrong drivers, will lead to the network card not working properly.
Troubleshooting: In the device manager window, check the network card options to see if the driver is installed normally. If the network card model is marked with "!" . Or "X" means that the network card can't work normally at this time. The solution is simple, just find the right driver and reinstall it.
2. There are conflicts between network card devices. The network card device conflicts with other host devices, which will cause the network card to fail to work.
Troubleshooting: Most disks are attached with programs to test and set the parameters of network cards, and check the parameters such as connector type, IRQ, I/ O port address, etc. If there is a conflict, as long as the network is reset (some jumpers must be adjusted), or the network card slot is replaced, so that the host thinks that it is a new device to redistribute system resource parameters, the network can generally return to normal.
3. The network address parameter of the host is set incorrectly. Improper setting of host network address parameters is a common host logic fault. For example, the IP address configured by a host conflicts with other hosts, or the IP address is not within the network range at all, which will cause the host to be unable to connect.
Troubleshooting method: check the connection properties window in the network neighborhood properties to see whether the TCP/IP option parameters meet the requirements, including IP address, subnet mask, gateway and DNS parameters, and repair them.
4. The host network protocol or service is not installed correctly. Improper installation of host network protocols or services can also lead to poor network connection. The protocol installed by the host must be consistent with other hosts on the network, otherwise there will be protocol mismatch and normal communication will not be possible. There are also some services, such as "file and printer enjoy service", which will prevent it from sharing resources with other users, and "network client service", which will prevent it from accessing resources provided by other users on the network. Another example is that the mail server is not properly set, which leads to the inability to send and receive mail, or the domain name server is not properly set, which leads to the inability to resolve the domain name.
Troubleshooting method: Check whether the installed protocols are consistent with other hosts in the properties of online neighbors (Windows98 system) or local connection properties window, such as TCP/IP protocol, NetBEUI protocol, IPX/SPX compatible protocol, etc. Secondly, check whether the service program corresponding to the service provided by the host is installed. If it is not installed or selected, please pay attention to installation and selection. Please note that sometimes you need to restart your computer for the service to work properly.
5. Host security failed. Another possibility of host failure is host security failure. It usually includes the theft of host resources, the control of hosts by hackers, the instability of host systems, etc.
Troubleshooting: The host resource is stolen, and the host does not control services such as finger, RPC and rlogin. Attackers can attack the host through the normal services or vulnerabilities of these processes, and even get administrator rights, and then have the right to copy and modify all the contents of the disk at will. It should also be noted that you should not easily enjoy the local hard disk, because this will lead to malicious attackers illegally using the resources of the host.
The host is controlled by hackers, which will lead to the host being out of the control of the operator. It is usually caused by the host being installed with a backdoor program. It is generally difficult to find such a fault. In general, possible vulnerabilities can be prevented by monitoring host traffic, scanning host ports and services, installing firewalls and patching systems.
The instability of the host system is often caused by malicious attacks by hackers or virus infection of the host. Kill the virus by antivirus software, and eliminate the possibility of the virus. Or reinstall the operating system, and install the latest operating system patches and firewalls, anti-hacking software and services to prevent malicious attacks caused by possible vulnerabilities.
Third, the conclusion
With the rapid development of computer network technology, network failures are also very complicated. The above summarizes several common faults and their troubleshooting methods. Generally speaking, we follow the principle of "soft before hard" for specific diagnostic techniques, but if we want to analyze specific situations, you need to accumulate these experiences for a long time. If you are a network administrator, you should also pay attention to the following aspects in network maintenance:
First of all, establish a complete networking document for query during maintenance. Such as system demand analysis report, general idea and scheme of network design, network topology planning, network equipment and network cable selection, network wiring, network IP distribution, network equipment distribution, etc.
Second, do a good job of network maintenance logs, especially some faults with low probability but great harm and some faults with high probability, and do a good job of complete maintenance documents for each machine to facilitate the investigation of faults in the future. This is also an accumulation of experience.
Third, improve the awareness of network security, improve the reliability of passwords, and install the latest operating system patches, firewalls, anti-hacking programs, etc. Prevent possible vulnerabilities for the host.