A typical case of router fault resolution

Router failure

One: Overwhelmed, the router outside the net port closes

1. Network environment

A unit uses Cisco router, leased telecom 30MB to do local access and L0MB Education Network dual-line internet, two years of stable network operation, routers also did not fail. As the number of network users increased, the original telecom 30MB can not meet the needs, and decided to rent telecom 1OOMB to solve the bandwidth problem. Telecommunications using optical fiber access to the unit room, the use of hundred trillion photoelectric converter after the conversion through the twisted pair of cables to the router outside the network, the router is using Gigabit electric port as an external network, because the photoelectric converter only 1O0MB, the port after the speed display 100MB.

2, the external network port flow is zero

After a few days of operation, the administrator found that every day when the router outside the network port traffic exceeded 50MBPS/S, the port will appear "receive Errors", traffic is very large, error message a lot. Suddenly one day, the external network can not be on, telnet to the router above, found that the corresponding external network of telecommunications no traffic, the display status of up, the other ports on the router work properly. The first reflection is the telecommunications side of the problem, is the telephone to inform the telecommunications over there, the other side quickly replied that there is no problem, and asked whether the photoelectric converter crashed. So the administrator will reset the photoelectric conversion, the fault is still. No way, had to restart the router, troubleshooting. Who knows, after less than one hours, the fault reappeared. After telnet to the router to perform shutdown and undo shutdown, troubleshoot the external network port. Who knows, to apply all the virus-related security policies to this port, modify the TCP MSS to 2048 (vendor default 1460), and the failure still occurs.

3. Fault Analysis

The administrator found that when the fault occurred, the CPU display 23%,memory is 33%, not too high, the key is that other interfaces are working properly, it appears that the problem is still on the top of this port. But this port has been used for two years, upgrade expansion before the port does not appear normal communication, port hardware should be what the problem. Through the network management software to the port before the closure of the flow detection, found that the port before the closure of a large flow through (over 80mbps/s), the display port error message is also more. Through the analysis that should be the network traffic is too large, the utilization rate caused by too high. The port is not normal after the traffic exceeds 80%. If the port works in Gigabit mode, the 100MB bandwidth uses only that port 10%, so the port can be handled easily.

4. Solution

More Wonderful content: http://www.bianceng.cnhttp://www.bianceng.cn/Network/lyjs/

After finding the sticking point, the recommended solution is to buy gigabit photovoltaic converters instead of the original hundred devices, and the price is relatively cheap. However, in order to ensure the stability of the network operation, the unit decided to purchase a Gigabit optical port routing module directly using optical fiber for communication and reduce network delay. Telecommunications is controlled by the port speed limit to ensure the provision of hundred gigabit bandwidth. Run through a period of time, found that the port in addition to a small number of error messages, no longer have the port shutdown without reason.

Router failure

Second: Why the router failed to contract

In the router configuration process, often encountered such a problem: network communication is normal, routers can successfully route packets to the target network, but the packet sent from the router failed, failure performance for the router ping target network failure, the following is a typical case.

(1). Description of the phenomenon

When a network configuration is complete, an administrator discovers that when a ping is sent from a PC (6.159.245.195) to the target network (6.159.245.65/26), the router R1 can successfully forward the packet, but from R1 to the target network ( 6.159.245.65/26) Ping failed while sending a ping.

(2). Error-Arranging process

First, track the path that the ping passes through. Check the R1 routing table, and the destination address 6.159.245.65 can match 0.0.0.0/0 in the routing table. Check the routing tables for R2, R3, and R4 to find the routing table entries that match the destination address.

Then, trace the path that the ICMP response response packet passed through. To complete this step, to explicitly respond to the source address of the packet, when the PC sends a ping, the response Reply packet's destination address is 6.159.245.195. When a router R1 a ping, the target address of the response reply packet is 71.170.0.146. The routing table entries that match 6.159.245.195 are found against the R4 routing table, and no routing table entries that match the destination address 71.170.0.146 are found. It appears that the ICMP response reply packet was discarded at R4 processing, and PMG failed when the ping was sent from R1 to the target network R4 (6.159.245.65/26).

The workaround is to add a static route to the 71.170.0.144/30 on the router R4, and the next hop address is 71.170.0.214. When you're done, when R1 sends a ping to R4, it turns out that everything is fine.

Router Troubleshooting Experience Summary: Although this kind of network failure will not affect the normal communication of the network, the process is also very simple, but in the network failure analysis and exclusion, we have to consider the complete communication process.