Classification and troubleshooting of common network faults

In the current network management system, due to the diversity and complexity of network fault, network fault classification methods are different. According to the nature of network fault, it can be divided into physical fault and logic fault, and it can be divided into line fault, router fault and host fault according to the object of network fault.

One, according to the Nature of Network fault division

1. Physical failure

Physical failure, refers to equipment or line damage, plug loose, the line is seriously electromagnetic interference and so on. For example, a network of a line suddenly interrupted, such as the installation of network monitoring software can be detected from the monitoring interface of the line suddenly fell down or the system pop-up alarm interface, more directly reflected in the line port on the Radio management Information system can not be used.

Solution: first with the DOS command set ping command to check the line and network Management Center Server port connectivity, if not connected, check whether the port plug is loose, if loose is inserted tight, and then use ping to check, if the general rule of failure to solve. It is also possible that the line away from the network Management center of the end of the plug loose, you need to check the terminal equipment connection status. If the socket is not a problem, you can use the network cable test equipment for path testing, found that the problem should be replaced by a network cable.

Another common physical failure is the network plug mistakenly connected. This situation is often caused by no understanding of the network plug specification or the network topology.

Solution: Familiar with the network plug specifications, such as t568a and t568b, to find out the color and significance of each line in the cable, make the plug to meet the specifications. There is also a situation, such as two routers direct connection, should let a router's exit to connect another router's entrance, and this router's entrance connects another router's exit, then the production of the network cable should meet this feature, otherwise it will lead to network misunderstanding. But like this network connection fault appears to be very covert, to diagnose this kind of fault has no special good tools, only rely on the experience of network management to solve.

2. Logical failure

A common scenario in a logical failure is a configuration error, which refers to a network exception or fault that is caused by a network device's configuration. The configuration error may be incorrect setting of router port parameters, or router routing configuration errors such as routing loops or the remote address is not found, or the netmask is set incorrectly. For example, the same network of a line failure, found that the line has no traffic, but can ping the port at both ends of the line, it is likely that the routing configuration error caused the loop.

WORKAROUND: Diagnose the failure with the Traceroute tool, you can find two IP address loops after a paragraph in the traceroute result. At this time, is generally the end of the line to route the port and point to the line near, resulting in IP packets on the line back and forth repeatedly. At this point you need to change the remote router port configuration, set the route to the correct configuration, you can restore the line. Of course all actions to handle the fault are recorded in the log to prevent recurrence.

Another type of fault in a logical failure is the shutdown of some important processes or ports, and the overload of the system. For example, the router's SNMP process accidentally shuts down or dies, then the network management system will not be able to collect any data from the router, so the network management system has lost control of the router. Also, the line is interrupted, there is no traffic, then Ping found the line near the port Ping impassability.

Workaround: Check that the port is down, which means the port has been shut down, causing the failure. The connection can be restored by simply restarting the port. In addition, a common situation is that the router is too overloaded, the performance is the router CPU temperature is too high, CPU utilization is too high, and memory margin is too small, although this kind of fault can not directly affect the network connectivity, but it affects the quality of the network to provide services, but also easily lead to the damage of hardware equipment.