Hi,
There are two possible causes:
1. The SNMP NMS frequently queries packets. As a result, the number of packets sent to the CPU exceeds the CPCAR threshold, the excess packets are discarded.
2. If the current CPU usage is low, increase the CPCAR value of the SNMP protocol. The default CPCAR value is 128. Change the CPCAR value to 256. If the alarm persists, increase the CPCAR value based on the CPU usage.
If no attack defense policy is configured on the switch, run the following command:
[Huawei]cpu-defend-policy test
[Huawei-cpu-defend-policy-test]car packet-type snmp cir 256
[Huawei-cpu-defend-policy-test]q
[Huawei]cpu-defend-policy test
[Huawei]cpu-defend-policy test global
If an attack defense policy has been applied, add car packet-type snmp cir 256 to the policy.

https://support.huawei.com/hedex/hdx.do?docid=EDOC1100126530&id=EN-US_LOGREF_0177122986&lang=en