Troubleshooting in networks - Trend Communications Ltd.

Troubleshooting in networks ‐ but how???The increasing complexity of the LAN environmentToday's LAN networks are becoming increasingly complex and fast ‐ and at the same time moresusceptible to error. While just a few years ago, a LAN normally consisted of just 10 or 100Mbit/sEthernet computer to server or server to server connections, nowadays there is a broad range ofdifferent transmission channels, speeds and devices. Firstly, at 1 gigabit/s a modern LAN has becomeconsiderably faster; secondly, it consists of a lot more than just the “classic” Ethernet connection viacopper. The transmission media today are no longer copper but glass fibre and wireless connections ‐with the term “wireless” not just referring to the common WiFi but also to media such as directionalradio or laser. And a company's LAN – “local area network” ‐ is no longer as “local” as it used to bebut can spread far over different continents thanks to technologies such as VPN (Virtual PrivateNetwork). A computer can access servers at different company sites all over the world via theEthernet connection within the company network, for example.If errors occur, these may be located in a wide range of different media and layers of the network.Errors in passive wiringThe most common network errors are still those at the physical level, i.e. problems with thetransmission medium such as classic wiring problems, problems with glass fibre cables or receptionproblems in WiFi networks.In attempting to locate the error, a simple wiring tester and a laptop with a few commands such asPing is no longer sufficient.So the user of a network should consider which media are operated within the network in order toselect the appropriate troubleshooting device.Troubleshooting devices have to be able to offer very good testing facilities at the physical level aswell as providing varied diagnosis options on the active LAN side.Figure 1:Display of a split pair error on wires 4‐5 and 7‐8 as part of a wiring testOn the physical copper level, devices have to be equipped with excellent error search functions forwiring. They should be capable of locating errors along a link and links within a panel so as to be ableto provide the user with fast and clear assistance in eliminating the error. Here it is especiallyimportant that the device finds errors in each individual wire and not ‐ as is common in many lowcostdevices ‐ errors in wiring pairs. Split pair errors should also be reliably detected. The split pairerror consists of identical wire exchanges on both sides of the link, for example wires 2 and 3 on bothsides of the link. So all eight wires are connected 1:1, but the Ethernet signals no longer run in pairs(pairs 12, 36, 54 and 78) but in split wires (in the above example 13, 26, 54 and 78), which results inan extremely high level of interference due to cross‐talk between the pairs. The link may still beestablished, but data transmission is extremely slow or fails completely due to the high level ofinterference. In practice, such split pair errors can easily occur, for example if two white wires areswapped by mistake. If the measuring device is unable to detect this error, the trouble‐shooter has

the impression that data transmission is slow and the error is on the active side, e.g. an overloadedswitch or a slow internet connection.On the glass fibre side, the device should be sufficiently versatile to be able to handle all commonwavelengths and transmission standards. Here, a fundamental distinction has to be drawn betweenmeasuring the attenuation of the passive link with a light source and a power meter or going straightto the active side. Some troubleshooting devices on the market have an SFP interface which can beused via commonly available SFP modules. The diagnosis options on the passive side are somewhatlimited in this case, but the benefit is that you can immediately use the SFP modules which match thesystem to connect the device via optical fibres.Diagnostics on EthernetThe diagnostic functions on the active LAN side are equally important. Once you have clearlyestablished that the passive wiring is fine, it might be necessary to continue the error search on theactive side. Troubleshooting devices should be compact and easy to operate. Compact, because thesearch for errors can often not be carried out from one place ‐ to be absolutely sure, the user has totest at various points in the network. A laptop or a large, unwieldy device is not much help here. Easyto operate, because the Ethernet world can quickly get very complex and it is easy to lose track. Aclear, well structured menu helps the user focus on the essentials.Essentially, there is a distinction between devices with one port which are connected to a LAN as anendpoint or devices with two ports which can additionally be connected to diagnose between twoEthernet devices in inline mode.EndpointThe endpoint mode is important to quickly obtain an overview of an Ethernet network.Troubleshooting devices can test Ethernet speed at the connected port in this mode and showinformation on the IP address and DNS (Domain Name Server) in use, for example.Network mapThis can be used to run a simple review of all participants in a company network. Clever devices cansave this review ‐ or network map, as it is called. The user can then reload the map later andcompare the network with this older map so as to find changes to the network more easily. Forexample, new participants or devices on the network can be found quickly, or it is possible to seewhich devices are missing. Here is an example: The user can no longer access a particular server onthe network. The network map can be used to quickly verify whether the server is still accessible onthe network. If the troubleshooting device also has an integrated trace‐route function, all stationsalong the communication path can be listed. In this way, the user can quickly establish at which pointin the network the problem is located.Figure 2:How many devices are active on the company network? A network map provides the answer.

InlineThe inline mode is of interest if you want to run diagnostics on a specific device or if you areinterested in data flow between two points in the network.Troubleshooting devices can now show the type of Ethernet connection at the two ports andtherefore identify a speed mismatch (e.g. one device is set to 100 FD and the other to 1000 FD). Itshould also be possible to investigate the data stream on this line more precisely. A goodtroubleshooting device should be in a position to generate so‐called top ten lists. These top ten listswill usually contain the ten addresses which send and receive the largest volume of information, theten most used protocols and the ten addresses which most frequently generate errors. If the internetconnection in a company is felt to be especially slow, for example, this type of troubleshooting devicecan be connected between the internet router and network. All internet traffic will then run throughthe troubleshooting device. Firstly, this allows the actual data transfer rate to the internet to bemeasured; secondly, the top ten lists can help detect whether certain devices or participants in thenetwork are using a particularly large amount of bandwidth.For example, if a PC is very slow on the network, the troubleshooting device can be connectedbetween the PC and the switch so as to observe the data traffic of this specific machine. It is thenpossible to tell whether this machine is generating too many errors or if an unwanted downloadmight be running in the background which is slowing down the network connection.Figure 3:The quality of VoIP connections is simple to check in inline modeVoIPIf the troubleshooting device also has diagnostic options for VoIP (Voice over IP) connections, it canbe connected between a VoIP telephone and the network so as to identify any potential problemswhen speaking on the phone. For example, if voice quality is poor or if calls are constantlyinterrupted, the troubleshooting device can use special VoIP measurement parameters such as jitterto provide an important indication of the cause of error.WiFiAs mentioned at the beginning of this article, WiFi is a fixed element of many company networkstoday. “Fixed” in the sense of indispensable, but not in the sense of stable. WiFi networks aresusceptible to error due to their wireless technology. For example, a typical situation in a companyhoused in a multi‐storey building is that the range of the WiFi network is suddenly reduced, from oneday to the next. No cause of error can be found, the AP (access point) seems to work since there isgood reception near the AP and communication works. Other APs are not identified by the laptopused for diagnostic purposes. What the IT technician is not aware of is that an AP has been installedby the neighbouring company on the next floor which is unfortunately using the same channel. Sincethis AP is “hidden”, i.e. it does not transmit an SSID, it is initially invisible to other devices on thenetwork, so the laptop does not locate it in its search for wireless networks. Without a propertroubleshooting device, the user is stuck and has no chance of finding the cause of the error. In

practice, this often results in more access points being set up, which only has the effect of reducingthe range of each access point even further.Another common error is that neighbouring APs use neighbouring channels for data transfer. Manyusers are not aware that this also heavily reduces the effective range of both APs.Finally, one AP is connected by means of a copper wire connection. This connection may also bedefective. Once again, a troubleshooting device can be connected inline between the AP and thenetwork so as to locate any transmission or configuration errors in the AP.Figure 4:With the LanXPLORER network testers from IDEAL INDUSTRIES, network troubleshooting is a pieceof cake!SummaryThe days of pure 10/100Mbit or Fast Ethernet copper networks are over. Nowadays, networks aremade up of varying transmission speeds and media with a wide range of services. This variety isimportant since the LAN has to perform a wide range of functions, but on the other hand it makesthe network more susceptible to disturbances. Smooth operation of networks is more importantthan ever, since downtime has to be avoided at all costs. This is why it is important to usetroubleshooting devices which support the versatility of modern networks and provide the user withswift help in the event of an emergency.www.idealindustries.co.ukAuthor:Dipl.‐Ing. Konstantin HüdepohlMarket Development Manager, IDEAL INDUSTRIES GmbH