Five strokes to resolve the security speed of network failure

LAN in the long run, there will always be some strange network fault, as a network administrator of the most important task is to locate the network fault node in time, and find the cause of the failure, and find ways to quickly solve it. Normally, when the Web site is not accessible, network connectivity testing, such as failure phenomenon, these failures because all have a clear phenomenon, we can quickly locate the failure node according to these phenomena, failure to find the cause is not very troublesome. However, there is a type of network failure is particularly annoying, clear network connectivity is smooth, but the speed of Internet connection is slow, that is what factors make the network access speed so slow? In fact, the slow speed of network access factors than more, we only have to troubleshoot each, in order to let the network access full speed Connection!





excluding cable barriers





network access from the physical cable to the beginning of the connection, do not think you are using off-the-shelf cable, there will be no problem. In fact, about half of the network failures are connected to the physical cabling, the production of network cable and the method of Plug and pull are closely related to the performance of network transmission. For example, if the physical connection cable we use is not manufactured according to the T568A Standard and the T568B standard, then the Internet signal is transmitted on the cable , it is easy to be disturbed by the outside world, when using this cable to access the Internet, it is easy to encounter the phenomenon of the network speed is unpleasant. In addition, if the network cable is not in solid contact, Internet access will also encounter intermittent symptoms.





so how to make sure that the Internet slow fault, is caused by the physical cable? Many times, in order to determine whether a network cable is a problem, we will insert the cable into a working laptop computer, on the Internet to test, if the test results are normal, the basic confirmation of the network cable is not a problem; in fact, this is not necessarily accurate, sometimes a network cable that is not made according to the rules , connected to the notebook computer all normal, but connected to the normal desktop computer, it will appear just started surfing the normal speed, and then very slow, and even can not normal internet phenomenon, the reason is this phenomenon, mainly notebook computer with the network card equipment quality is high, and the desktop computer installed network card device quality is significantly inferior to the notebook computer.





Therefore, when the network cable barrier is excluded, we first want to confirm that network cable must meet the T568A standard and T568B standard, for the network cable that does not conform to the production standard should be resolutely discarded, in order to prevent the internet to produce recessive trouble in the future; second, with professional tools to test cable connectivity, if you don't have a professional testing tool on hand, You can use a notebook computer combined with a desktop computer for a comprehensive test, confirm the connectivity of the network cable is normal; third, check whether the network cable length within 100 meters, for those too long distance of the physical cable should be as short as possible to prevent the Internet users at any time will be too long cable coil around, affect the normal transmission of the internet signal.





Remove Port Barriers





hubs, switches, routers and other devices, LAN or WAN environment is an important equipment, the port of these devices under the impact of large-capacity data information, it is easy to appear a variety of fault phenomena, such as the port of suspended animation, lock dead phenomenon, the port was down and so on, once these important ports appear unexpectedly, Then the LAN will have a large number of client systems can not be normal access to the Internet, or access to the network is not fast.





to this end, when there are many computers in the local area network at the same time the slow speed of the fault phenomenon, we need to the important network equipment port work status check; For example, you can log into the backend management system of these devices, use the show command or display command to scan, Diagnose the working status of each port; If the fault node is not found in this way, we can try to use the Professional network management program to view the data traffic of the hubs, switches, routers and other important ports, and then observe their state changes. If the data flow of these ports has been very large, it is not only the problem of port bottlenecks, it is likely that there are similar network loops, broadcast storms and other anomalies in the LAN, on the contrary, if the data flow of these important ports is only at the peak period of access is relatively large, and the normal flow can be restored to the regular state, Then we can basically assume that these ports may be the network data flow bottleneck position, at this time we should find ways to increase the bandwidth capacity of the port. We can use a variety of methods to achieve this, such as the division of multiple VLANs, you can adjust the port settings, you can upgrade network devices, you can install multiple network adapters, and so on, to increase bandwidth capacity size.





Eliminates loop obstacle





In the LAN environment with small network size, due to the relatively small number of network nodes, the network structure is not particularly complex, the probability of the phenomenon is often small. However, in the environment of large group network and complex network structure, often reserve some spare network lines, network administrators a little attention, when they are connected to the network, it may cause network loop phenomenon; In addition, there may be a connection loop phenomenon when multiple switches or hubs are stacked to connect. This is because the uplink port of a switch or hub often shares an identical transport channel with the first Ethernet port, which causes a loop failure when we accidentally connect the uplink port with the other Ethernet port to the same client. For example, from the LAN computer room leads to a network cable, connected to the second floor of the 213 rooms, and then from 213 rooms to 214 rooms, if we then directly from the computer room to draw a network connection to 214 rooms, when the LAN room, 213 rooms, 214 rooms between the network lines, May constitute the network loop, then the Internet packet will be non-stop to send and verify operations, eventually resulting in the overall network transmission speed is relatively slow.





in order to avoid the occurrence of network loop phenomenon, we are in the laying of network cable, we must develop a good operating habits: Each network cable must be affixed to the obvious label, can not use the standby line where possible, even if the use of standby cable, it must be good record description. When encountered network access slow, suspected network loop fault phenomenon, we can first look at the standby cable connection is correct, in the confirmation of the standby cable connection is correct, and then use the partition, segmentation method, to step through the investigation.





Remove virus disorder





now the virus in the Internet network is rampant, and the harm is very large, once the LAN hit the network virus attack, the entire network of Internet speed may be greatly reduced, and may even cause LAN paralysis failure. The current LAN is very easy to encounter ARP virus attack, after poisoning may have one or more computers can not access the Internet, even a large number of computers can not access the Internet.





In the removal of ARP virus barriers, we can use the LAN switch management functions, to quickly locate infected computers; For example, for H3C-branded switches, we can go into the switch backend management system and execute string command "Dis dia" to scan each switch port in the LAN. After scanning, we can know which virtual subnet of the computer has ARP virus, and we can also look at the virus infected with ARP network card Physical address, then we can combine the data, find the ARP virus infected with the specific computer, the computer from the network isolated, At the same time using the latest version of anti-virus software to its comprehensive, completely virus killing operations, wait until the virus clean, and then reconnect it to the local area network, I believe that the LAN network access speed can be restored to normal.





of course, in order to avoid the network virus to drag down the speed of the Internet, we have to do before the computer access to the network, to do a full security protection work; For example, install hardware firewalls, network anti-virus software in the LAN, the client operating system to upgrade in a timely manner, to regularly install a variety of vulnerability patches, Service shutdown run, and so on.





rule out Storm disorder





after excluding some of the above factors, if still unable to improve the speed of network access, we can try to check the network storm phenomenon in the LAN, because this phenomenon is particularly easy to happen, such as as long as the network equipment in the LAN hardware damage, there may be network storm phenomenon, This phenomenon often has a certain concealment, many people tend to ignore the existence of this phenomenon. Generally speaking, as the number of workstations in the LAN is increasing, the number of broadcast packets sent to the network will increase, once the value exceeds 30% of the network traffic, then the Internet users will obviously feel that the network transmission speed will slow down. The most likely cause of the network storm is the network equipment hardware damage, it will continue to send a large number of broadcast packets to the network, thus triggering the broadcast storm phenomenon, until the communication of the paralysis phenomenon.





to this end, when the internet is found that the speed is not normal, we can troubleshoot the local area network equipment hardware damage occurred. In doing this kind of troubleshooting, we can first take advantage of the management function of the switch to find the port of traffic anomaly, followed by the replacement device or replace the port method, followed by the troubleshooting. In general, hubs, switches, network cards and other devices are most likely to occur hardware damage.