Troubleshooting Service intermittence based on the network of MSTP&VRRP

Service intermittence occurs based on the network of MSTP&VRRP.

Possible Causes

1) A switch receives massive TC packets and frequently clear address tables, which causes network flapping.

2) The CPU usage on a switch is too high, which causes that protocol packets cannot be processed in time and then causes network flapping.

3) BPDU packets sent to the CPU of a switch are lost, which causes network flapping.

4) VRRP packets sent to the CPU of a switch are lost, which causes network flapping.

Troubleshooting ProcedureStep1:Check whether a switch receives massive TC packets.

Run the show logging command tocheck whether the switch receives massive TC packets.

--->>If massive TC packets are received in network flapping, the network flapping may be caused by the TC packets. You need to configure TC protection by using the spanning-treetc-protection command on the switch and identify the source of the TCpackets.

Step2: Check whether the CPU usage of a switch is toohigh.

--->> If the CPU usage exceeds 90%, the network flapping may be caused by the high CPU usage. Because of that corresponding protocol packets cannot be processed in time and then causes network flapping.You need to check the processes that cause the high CPU usage.

Step3: Check whether BPDU packets sent to the CPU of a switchare lost.

(1) Check whether the BPDU packets sent to the CPU of themanagement board are lost.

--->> If this value is greater than 0 and increases, it indicates that BPDU packets sent to the CPU of the management board are lost, which may be the cause for network flapping.

(2) Check whether the BPDU packets sent to the CPU of the line card are lost.

--->> If this value is greater than 0 andincreases, it indicates that BPDU packets sent to the CPU of the line card arel ost, which may be the cause for network oscillation.

If the BPDU packets sent to the CPU are lost, it is recommended to set the CPP threshold for BPDU packets to a larger value. In addition, you need to check whether the number of BPDU packets in the environment meets the expectation.

Step4: Check whether VRRP packets sent to the CPU of aswitch are lost.

(1) Check whether the VRRP packets sent to the CPU of the management boardare lost.

--->> If this value is greater than 0 and increases, it indicates that VRRP packets sent to the CPU of the managementboard are lost, which may be the cause for network flapping.

(2) Check whether the VRRP packets sent to the CPU of the line card arelost.

--->> If this value is greater than 0 andincreases, it indicates that VRRP packets sent to the CPU of the line card arelost, which may be the cause for network flapping.

If the VRRP packets sent to the CPU are lost, it is recommended to set the CPP threshold for VRRP packets to a larger value. In addition, you need to check whether the number of VRRP packets in the environment meet the expectation.

Step5:Fault Information Collection

Collect log information (pay attention to the time stamp and the time accuracy) for background analysis.

show logging

showcpu

showcpu-protect mb

show cpu-protect slot X //X indicates the slot ID, such as use this command onSlot 4 of the stand-alone Switch8610: show cpu-pro slot 4;on Slot 4 of Chassis2in the VSU: show cpu-pro slot 2/4.