The problem of network interruption when the user equipment is running

Problem Description

A brief description of the network:

LAN--USG5120BSR (G0/0/0)--corotator (optical fiber)---WAN

Phenomenon and problem description:

1. The network used USG2160 before, and it was normal. After the USG2160 is replaced by usg5120BSR, it is often unable to access the Internet. The external network is connected to the G0/0/0 of the USG5120BSR through the co-rotor optical to electrical power conversion, the system is 100M, full duplex, and the G0/0/0 physical protocol is used when the network cannot be accessed. They are all UP, and learned the ARP of the opposite end, but cannot ping the opposite end IP, and can only restore the network through the shutdown undoshutdown interface, or plug and unplug the network cable 2. Confirm with the customer that the G0/0/
0 port uses the self Negotiation, the above problem occurs, after forcing the duplex rate, the problem remains.

warning message

none

process

Related process:

1. Quickly restore the user's business, the specific operation:

Delete the following configurations under the GE0/0/0 interface, and use auto-negotiation for the rate and working mode.
 speed 100
 duplex full
2. Check that there is no problem with the configuration of the user equipment

3. The version is too low, upgrade to the latest version (v100r005c00spc500), but the problem remains.

4. Message analysis, by remotely obtaining messages, analyze the messages of the device.

After analyzing the message information and log information, there are only messages sent from G0/0/0, and no messages are received. However, since the messages are obtained remotely, they belong to software acquisition messages, which can only prove that the The message has been sent out. As for whether the underlying driver and physical port have sent the packet, it cannot be confirmed.
5. To analyze the message of the underlying driver and physical port, the steps are as follows:

(1). So when there is a problem, please clear the interface packet statistics

reset counters interface

(2). Execute the Ping gateway

(3). Check the interface statistics again

dis interface GigabitEthernet 


(4). After executing these, enter the hidden view to see the statistical information of the data plane

[USG2200-hidecmd]display dataplane statistic module
(5). Clear data plane statistics

[USG2200-hidecmd]reset dataplane statistic module                              
20:10:31  2008/01/06                                                           
  will reset normal packet statistic! Continue?[Y/N]:y                         
[USG2200-hidecmd]

(6) Execute the Ping gateway

Look at the data plane statistics again

USG2200-hidecmd]reset dataplane statistic module 

Finally, collect interface packet statistics again.

6. Through analysis: the possibility of the problem of our equipment is relatively small. It should be caused by the ARP entry of the switch directly connected to our equipment not being refreshed correctly or there are other same ip addresses in the network

7. After contacting with the operator's engineer, the test found that it was caused by the conflict of the public network address. After modifying the public network IP address, the problem no longer appeared.

Root cause

problem causes:

1. User device configuration problem

2. The user device version is too low

3. Problems with the other party’s equipment

Guess you like

Origin blog.csdn.net/lnvjp/article/details/132332781