Failure to access the Internet due to fiber Signal Attenuation

One day, an online user in a room called to reflect a problem with the Internet. At the beginning, the author thought this was caused by the Internet user, so he was advised to change an IP address, or try re-plugging the network cable. unexpectedly, the phone has not been hung up, and other online users have called in succession to concentrate on the failure of the Internet, which makes the author feel the seriousness of the situation. Since a large number of users are unable to access the Internet, I have to go to the fault site to handle it.

Networking Environment

When I arrived at the fault site, I first made a careful investigation and found that the users who could not access the Internet basically came from the fourth floor. I checked the network structure topology, I found that all the users on the fourth floor are connected to the two sets of stacked Huashan 3050 switches through the cat5e network twisted pair. These switches are centrally placed in the dedicated wiring room on the fourth floor. To achieve high-speed connection with LAN networks, access switches on all floors use SFP/1310nm/10km modules and connect to core Switches of LAN through Gigabit multimode optical fiber, the core switch is the Huashan 8500 route switch, which is located in the central IDC of the LAN. The distance between the access switch on the fourth floor and the LAN core switch is about 200, the connection methods used between them are: access switches-optical fiber terminals-core switches, optical fiber cables and optical fiber terminals, and other devices are provided by local broadband operators. To effectively prevent the impact of broadcast storms on the operation of the LAN, all computers in the LAN are divided into eight VLANs based on their offices, the Gateways of each VLAN are uniformly set on the core Switches of the LAN. at ordinary times, all computers on each floor can access the Internet normally, and the Internet access speed is very fast.

Troubleshooting

Due to the failure of Internet users, all come from the 4th floor network, the author subconsciously believes that this fault is definitely caused by the 4th floor network itself; in order to determine whether your analysis is correct, I immediately used my laptop to test the connection to the networks on other floors. I found that accessing the Internet on other floors is normal. Obviously, the core switch of the LAN works normally, the problem seems to be carefully searched in the network on the fourth floor.

1. Check the vswitch

Since many users on the fourth floor of the network cannot access the Internet at the same time, the problem should be checked from the access switch on the floor; the author immediately rushed to the dedicated wiring room on the fourth floor, observe the signal light status of SFP/1310nm/10km module on the access switch and find that the signal light is off, which means that the switch cascade port on the fourth floor is not enabled properly, no wonder all computers connected to the vswitch on the fourth floor cannot access the Internet normally. Is there any problem with the access switch on the fourth floor? When there is a problem with a vswitch, regardless of the major or minor problem, we only need to restart the vswitch background system, and all kinds of failures will disappear immediately. This time, I have no exception, however, I am very disappointed that after the switch on the fourth floor restarts stably, all the computers on the fourth floor still cannot access the Internet, this indicates that the above fault may have nothing to do with the switch's own system.

2. Check the port

Is it possible that the cascade port working mode between the access switch on the fourth floor and the LAN core switch does not match? With this in mind, I think the possibility still exists. If their port working modes do not match, the communication between the fourth-floor network and the lan network will be abnormal, then, the network on the fourth floor will naturally fail to access the Internet normally. To check whether their connection ports match, I log on to the two vswitch backend systems as a system administrator and run the "interface" command to enter the view mode of the cascade ports, in this mode, run the "display interface xx" command (xx indicates the cascade port number) to view the configuration information of the cascade port and find that their working modes match, they all use 1000 m, full duplex mode.

3. query module

After removing the switching port, I plan to make some adjustments to the receiving and receiving ports of the optical fiber line. That is to say, on the Switching Module, I will swap the two inserted optical fiber core lines for a bit, insert the wire number inserted to the TX port to the RX port, and insert the wire number inserted to the RX port to the TX port. Because this operation needs to be performed at both ends of the core switch and the access switch at the same time, I ask another colleague to help at the core Switch site. What surprised me very much is that, when this colleague arrived at the core switch, he found the signal lights on the corresponding module. Why? Later, when we simultaneously switched the Port location of the optical fiber receiving and transmitting, we found that the signal lights on the module corresponding to the vswitch on the fourth floor were again lit, while the signal lights on the core switch module went off, will it be the connection to the fourth-floor network? After careful check, I confirm that this is not wrong with the optical fiber line, so why is the switch light on the fourth floor going on after the switch is sent and received, and the light on the core switch going off?

After careful analysis with colleagues, I believe that the switch module does not have any problems, because when the switch port is in the 1000 MB, full duplex mode, the optical ports on both sides do not need to be negotiated during communication. As long as the RX port of the Switching Module receives the optical signal, the signal light status of the corresponding port will be lit; on the contrary, if the ports of the vswitch are in the Automatic negotiation mode, the two core lines in the optical fiber cable must be connected at the same time to ensure that they are successfully negotiated during communication, the signal lights of the Switching Module on both sides should also be in the light state at the same time. However, the signal lights of one end switch module are in the light state, and the other end is in the off state, this indicates that there is no connection between the vswitch and the core vswitch on the fourth floor. After the receiving and receiving ports are redeployed, the signal lights of the switch module are still lit on one side and extinguished on the other side. Obviously, there is no problem with the switch module. The problem lies in the line of one core number.

4. Optical Fiber Query

Since the problem is that the fiber optic cable connecting the vswitch and the core switch on the fourth floor is located, what is the problem with the core line? I did not have any professional tools at hand, so I invited technicians from the local broadband operators to help with the fault detection and viewing. Technical staff of local broadband operators bring a tool such as optical power meter, which is used to determine whether the optical fiber cables are accessible or not. After repeated tests and inspections, the technicians found that there was a great signal attenuation in the middle of the optical fiber cable, resulting in one of the core lines could not receive the optical signal. Later, the technical staff used a professional tool called Optical Time Domain Reflector to test the signal sending and receiving of the optical fiber cable, and finally found the reason for the large attenuation of the optical signal, it turns out that the performance of the optical coupler has caused signal attenuation. After the new optical coupler device is used for replacement, the optical fiber cables are finally smooth, and all the computers on the fourth floor of the network have been tested, failure to access the Internet also disappears immediately.

Fault Summary

After careful troubleshooting, we finally found that the fault was due to fiber signal attenuation, resulting in a failure in accessing the Internet. In order to help you better solve the network failure caused by Attenuation of Optical Fiber signals, I have searched and summarized the causes of this phenomenon on the Internet, there are several main factors that may cause attenuation of Optical Fiber signals:

1. Jumper factors. In general, when laying optical fiber cables, the cables used can still be guaranteed in terms of performance. At the same time, during actual construction, we will also conduct quality tests on the Performance of optical fiber cables. If the quality is too low, we will find problems in the first place. However, the detailed factors of Optical Fiber Jumpers are often ignored, in addition, the quality of Optical Fiber jumpers on the market is also very different. Many of us choose Optical Fiber jumpers with an indifferent attitude. When we are not lucky, when a poor quality optical jumper is selected, the optical fiber signal can be reduced by 10 dbm, which is enough to cause network connection failure. To avoid interference from optical fiber jumpers, We must select jumpers from companies with hard quality and good reputation when establishing a network. Of course, if conditions permit, we 'd better use an optical power meter to test the patch cord we purchased at the purchase site.

2. Coupling factors. Because optical coupler is usually installed on the optical fiber distribution box, I believe many people will ignore its quality problems. However, in the process of actually managing the network, I found that if the optical coupler has poor performance, it may also cause a relatively large signal attenuation. Therefore, if you need to separately purchase the optical coupler during the networking process, you still need to purchase from companies with hard quality and good reputation.

3. coupling frequency factor. In order to connect optical fiber cables at different locations, we usually need to couple the optical fiber lines in the pipeline. In general, we perform coupling for each optical fiber line, the optical fiber signal degrades by about 1 dbm. If the coupling frequency is large, the attenuation of the optical fiber signal will be greatly increased, which will also affect the network connection speed.

4. Optical Fiber Connector factors. This is a common problem. If you pay close attention to it, you can find any pair of optical fiber lines, even well-communicated optical fiber lines, as long as the optical fiber connector is wiped several times repeatedly, the attenuation of the optical fiber signal can be reduced by about 2 dbm. Therefore, when the optical fiber line is working for a long time, or when the optical fiber line is in a dusty working environment, we should try to wipe the connector of the optical fiber cable. I believe this will not achieve the expected effect.

5. Exchange Module factors. Most of the time, multi-mode optical fiber lines are used in the pipeline or underground. However, when switching modules are used, the single-mode switching module is used, as a result, the connections between them may not match, which may cause serious data packet loss and increased signal attenuation. Similarly, if the Switching Module is multimode, however, the optical fiber jumper is single-mode, which may cause abnormal network connection faults. In addition, if the optical fiber line does not match the optical fiber jumper, it will also cause abnormal network faults.

Based on the above theoretical analysis, we can try to troubleshoot the optical fiber line connection failure by following the steps below:

First, use a professional optical power meter or Optical Time Domain Reflector to check whether the connectivity of optical fiber lines is normal. If optical signals can be sent and received normally, the connectivity of optical fiber cables is normal; if the optical fiber line cannot be normally received, there are two possibilities: one is that the two sides of the optical fiber cable do not match correctly, and the other is that the optical fiber line is broken, at this time, we must ask our professional Optical Fiber line engineers to check and solve the problem.

When the connectivity of optical fiber lines is confirmed to be normal, if the optical module port of the switch cannot be enabled properly, check the working mode and signal attenuation of the switch module, for specific troubleshooting steps, refer to the above sequence.