Switch failure in the operation is inevitable, but after the failure, it should be promptly handled, and the point of failure should be identified and remove as soon as possible, which is the responsibility of network management personnel. But for this, you must understand the type of switch failure and have the ability to analyze and process faults. To this end, this article gives a brief introduction to common types of failure in the switch and analyzed troubleshooting methods.

1. Power failure

External power supply’s instability, power line aging or lightning and other reasons lead to power damage or fan stop so that it cannot work properly. Or the damage to other parts of the machine caused by power supply will cause the problems of the switch.

If the POWER indicator on the switch panel is green, it indicates that it is normal. If the indicator is off, it indicates that the switch is not powered normally. This kind of problem is easy to find, also easy to solve, but also the easiest to prevent.

Solution: For this type of failure, you should first do the work of external power supply, generally through the introduction of independent power lines to provide an independent power supply, and adding a regulator to avoid instantaneous high voltage or low voltage. If the conditions permit, you can add UPS (uninterruptible power supply) to ensure the normal power supply switch, and some UPSs provide voltage regulator, and some do not, so you should pay attention when making a choice. Set up professional lightning protection measures in the room to avoid lightning on the switch damage. There are many professional companies for lightning protection projects, and they can be considered when it comes to the implementation of network cabling.

2. Port failure

This is the most common hardware failure, whether it is fiber port or RJ-45 port of twisted pair, you must be careful to plug the connector. If you accidentally smudge the fiber plug, it may cause the fiber port polluted which cannot communicate properly. If you are not careful when handling, it may also cause physical damage to the port. If you buy a too crystal head, it is easy to damage the port when inserting the switch. In addition, if the twisted pair on the port is exposed to the outside, if the cable is hit by lightning, it will cause the connected switch port to be damaged, or cause more unpredictable damage.

Solution: Under normal circumstances, the port failure is one or several ports damaged. Therefore, after excluding the failure of the computer connected to the port, you can replace the port to determine whether it is damaged. In the event of such a failure, you can clean the port with alcohol cotton balls after the power is turned off. If the port is indeed damaged, you can only replace the port.

3. Module failure

The switch is composed of many modules, such as stacking modules, management modules, and expansion modules. The chances of failure of these modules are small, but in the event of problems, they suffer huge economic losses. If you accidentally plug the module or carry the switch by the collision, or power supply is unstable, etc., it may lead to the occurrence of such failures.

Of course, the above-mentioned three modules have external interfaces, relatively easy to identify, and some can also be identified the fault through the indicator on the module. For example, the stack module has a flat trapezoidal port, or some switches have a USB-like interface. Management module has a C * OLE port used in the connection with network management computer to facilitate management. If the expansion module is a fiber connection, there will be a pair of fiber interfaces.

Solution: When removing such a fault, first ensure that the power supply of the switch and module is properly supplied, and then check whether each module is inserted in the correct position. Finally, check whether the cable connected to the module is normal. When connecting the management module, consider whether it uses the specified connection rate, whether there is parity, whether there are data flow control and other factors. When connecting an expansion module, you need to check whether the communication mode is matched, such as using full-duplex mode or half-duplex mode. Of course, if the module is confirmed to be faulty, there is only one solution, that is, you should immediately contact the supplier to replace.

4. Backplane failure

The modules of the switch are plugged into the backplane. If the environment is wet, the damp circuit board will in short circuit, or components due to high temperature, lighting, and other factors will cause damage to the circuit board and cause that it cannot work properly. For example, poor heat dissipation or too high environmental temperature will lead to increased temperature inside the machine, resulting in burned components.

Solution: In the case of normal power supplied by the external power supply, if the internal modules of the switch cannot work properly, it may be a bad backplane, and even if an electrical maintenance engineer has no idea. I am afraid that the only way is to replace the backplane.

5. Cable failure

In actual use, the cable failure often causes that the switch system or port does not work properly, so here such a failure is also into the switch hardware failure: connector plug is not tight, the cable is arranged in the wrong order or not standardized, the cable connection should use cross connection but use direct connection, in the cable two fiber use staggered connection, and the wrong line connection leads to network loop.

Solution: From the above several hardware failures, the poor engine room environment can easily lead to a variety of hardware failures, so  in the construction of the engine room, we must first do lightning protection and power supply, indoor temperature, indoor humidity, electromagnetic interference, anti-static and other environmental constructions, providing a good environment for the normal work of network equipment.

6. Improper configuration

Because beginners are not familiar with the switch, or the switch configuration is not the same, the administrator often has configuration error when configuring the switch. For example, that VLAN partition is not correct causes that the network is obstructed, port was falsely closed, switch and network card configuration does not match.

Solution: If you cannot ensure that the user's configuration is faulty, please restore the factory default configuration, and then make the configuration step by step. It is best to read the manual before configuration, which is one of the habits to be developed. Each switch has a detailed installation manual, and user manual, and there is a detailed explanation deep into each module. As a lot of switches manuals are written in English, so the users in poor English can consult the supplier's engineers and then do the specific configuration.

7. External factors

Due to the existence of hacker attacks, it is possible for the connected port of a host to send a large number of packets which do not meet the rules of the package, resulting in the too busy switch processor, too late packet forwarding, and then packet loss. There is also a situation that is the broadcast storm, which will not only take up a lot of network bandwidth but also a lot of CPU processing time. If the network is occupied by a large number of broadcast packets for a long time, the normal point to communication cannot normally work, and the network speed will slow down or be paralyzed.

Failure of a network card or a port is likely to trigger a broadcast storm. The switch can only split the collision domain without splitting the broadcast domain (in the case of no partition of VLAN). So the network’s transmission efficiency will significantly reduce when the number of broadcast packets accounts for 30% of the total communication.

Conclusion

As the phenomenon of switch failure is various, there is no fixed exclusion step. Some failures are often with a clear direction which one can identify. So we can make specific analysis only according to the specific circumstances. Of course, no matter what kind of failure is a difficult thing for a new network administrator, so if you want to be able to become a removal master on switch failure, you must accumulate experience in the daily work, and review the root causes of the problem and the solution to every problem is handled so as to better complete the work of network management.