No relevant resource is found in the selected language.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>


To have a better experience, please upgrade your IE browser.


Light Packet Loss Occurs When the NE80E Is Pinged Because of Layer 2 Network Flapping

Publication Date:  2013-01-07 Views:  83 Downloads:  0
Issue Description

When the NC server pings the NE80E that functions as a gateway, 1.5% of the packets are lost.


Handling Process

Since the NE80E cannot collect statistics about the Reply messages with which it responds to the NC server, you cannot check whether the NE80E has replied to the Request messages completely and correctly by collecting traffic statistics on the NE80E. However, the traffic statistics show that the NE80E receives all the Request messages that the NC server sends in the ping process.

Attach a PC to the NE80E directly and use the PC to ping the NE80E. Packet loss still occurs. Capture the packets and find that the NE80E can receive all the Request messages that the PC sends. However, some of the Request messages are not replied to. Therefore, the packet loss occurs on the NE80E.

An analysis of the captured packets shows that the NE80E sends an ARP Request message after each packet loss (For details about the captured ICMP packets, see the following attachment). The packet loss on the NE80E is caused by the loss of ARP entries.

A check of the ARP entries on the NE80E shows that the ARP entries are refreshed frequently. The aging time remains 20. That is, the NE80E refreshes its ARP table at least every minute.

Check the logs stored in the log buffer and find no abnormality.

Check the logs stored in the hard disk and find that a large number of BPDUs are sent out from GE 9/0/23 (nearly once every 30 minutes). The log information is as follows.

Feb 25 2009 01:23:37 DongZhiMen_NE80E_1 %%01MSTP/6/RECEIVE_CISTTC(l): MSTP received BPDU with TC, instance 0, port name is GigabitEthernet9/0/23.

Therefore, the NE80E frequently refreshes its ARP entries because it frequently receives BPDUs.

Based on the log information, an interface on the S8016 attached to the NE80E alternates between Up and Down frequently, generating a large number of BPDUs.

After shutting down the interface that alternates between Up and Down frequently on the S8016, the ping operation recovers and the packet loss rate is 0.

Root Cause

NE80E frequently refreshes its ARP entries because it frequently receives BPDUs.



The event of the NE80E receiving BPDUs is recorded in a log with low priority. By default, no log is stored in the log buffer when the NE80E receives BPDUs. You are strongly advised to configure the interface through which the S8016 and the server are connected as an STP edge interface and enable BPDU protection on the S8016 globally to prevent unwanted BPDUs from being generated. NE80E process the ARP entries in either of the follow modes after receiving BDPUs: Fast: Deletes all ARP entries related to the interfaces in the STP instance and learns new ARP entries driven by traffic. Normal: Sets the aging time of the ARP entries to 0 and refreshes the ARP entries based on the refreshing mechanism for normal ARP aged entries. The Fast mode takes less time in STP refreshing and convergence than the Normal mode but the Fast mode causes frequent network flapping easily. The Normal mode is stable. Select either of the modes as needed. For the default mode on a router, see the command reference in the corresponding software version..