【Problem Description】
The switch is running now at its optimum. Now once we add more customer networks now , the switch starts to ping loss ( CPU utilization never reaching to 60% , which is “huawei” suggested optimum level ) with CPU utilizations notmore than 40 % to 42 %, that means its starts to fail once its CPU is 42%only.
【Problem Analysis】
suggest customer adjust the CPCAR value as follow to avoidpacket loss:
[Huawei] cpu-defend policy defend [Huawei-cpu-defend-policy-defend] auto-port-defend enable [Huawei-cpu-defend-policy-defend] auto-port-defend protocol arp-request threshold 1000 [Huawei-cpu-defend-policy-defend] auto-port-defend protocol arp-reply threshold 1000 [Huawei-cpu-defend-policy-defend] auto-port-defend protocol icmp threshold 1000 [Huawei-cpu-defend-policy-defend] car packet-type arp-request cir 512 cbs 288000 [Huawei-cpu-defend-policy-defend] car packet-type arp-reply cir 512 cbs 288000 [Huawei-cpu-defend-policy-defend] car packet-type arp-miss cir 512 cbs 288000 [Huawei-cpu-defend-policy-defend] car packet-type icmp cir 512 cbs 288000 [Huawei-cpu-defend-policy-defend] auto-defend enable [Huawei-cpu-defend-policy-defend] auto-defend alarm enable [Huawei-cpu-defend-policy-defend] auto-defend action deny # [Huawei] cpu-defend-policy defend [Huawei] cpu-defend-policy defend global |