CPU过载还在重启？这5个命令连元凶都找到了

号主：老杨丨11年资深网络工程师，更多网工提升干货，请关注公众号：网络工程师俱乐部

交换机CPU使用率过高是常见问题，可能引发响应缓慢、管理中断、甚至业务丢包。

但盲目重启不仅治标不治本，还可能掩盖真正的故障根源。

华为交换机的这五个诊断命令，能帮助你快速定位“元凶”

——是STP震荡？ARP泛洪？还是非法配置引发的协议风暴？

第一步：确认CPU使用率是否真的异常

使用 display cpu-usage 查看当前CPU占用情况。

<Huawei> display cpu-usage
cpu-usage : 75% &nbsp; ---- 最近5秒CPU利用率
cpu-usage (5 seconds) : 75%
cpu-usage (1 minute) : 68%
cpu-usage (5 minutes) : 65%

CPU utilization&nbsp;for&nbsp;five seconds: 75%: &nbsp;User: 15% &nbsp; System: 60% &nbsp; IRQ: 0%

📌&nbsp;解读：

• **5秒利用率 > 70%**：可能存在突发流量或协议震荡
• **1分钟/5分钟持续 > 50%**：需重点关注，存在持续性负载
• System 占比过高：通常是协议或中断处理占用，是排查重点

第二步：查看CPU占用“TOP进程”

——display cpu-usage process

这是最关键的命令，列出按CPU占用排序的进程。

<Huawei> display cpu-usage process
Get cpu usage successful.
-------------------------------------------------------------------------------
&nbsp; 1. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp; 2. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp; 3. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp; 4. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp; 5. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp; 6. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp; 7. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp; 8. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp; 9. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;10. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;11. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;12. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;13. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;14. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;15. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;16. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;17. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;18. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;19. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
&nbsp;20. 0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp; &nbsp;0% &nbsp;&nbsp;
-------------------------------------------------------------------------------
&nbsp; 1. CPU 0 &nbsp; 2. CPU 1 &nbsp; 3. CPU 2 &nbsp; 4. CPU 3 &nbsp; 5. CPU 4 &nbsp; 6. CPU 5 &nbsp; 7. CPU 6 &nbsp;&nbsp;
&nbsp; 8. CPU 7 &nbsp; 9. CPU 8 &nbsp;10. CPU 9 &nbsp;11. CPU 10 12. CPU 11 13. CPU 12 14. CPU 13 &nbsp;
&nbsp;15. CPU 14 16. CPU 15 17. CPU 16 18. CPU 17 19. CPU 18 20. CPU 19

Process&nbsp;in&nbsp;ascending order of CPU usage:
No. &nbsp; Process name &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; CPU% &nbsp; &nbsp; Runtime(ms)
1 &nbsp; &nbsp; OSPF_Extra &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; 45.2 &nbsp; &nbsp; 12345678
2 &nbsp; &nbsp; STP &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;20.1 &nbsp; &nbsp; 5678901
3 &nbsp; &nbsp; ARP &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;15.3 &nbsp; &nbsp; 4567890
4 &nbsp; &nbsp; SNMP &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; 2.1 &nbsp; &nbsp; &nbsp;123456

📌&nbsp;解读：

• **OSPF_Extra 占 45.2%**：可能是OSPF邻居频繁震荡，触发大量SPF计算
• **STP 占 20.1%**：端口频繁UP/DOWN，导致拓扑重计算
• **ARP 占 15.3%**：可能存在ARP请求泛洪或攻击

第三步：检查接口状态变化

——display trapbuffer

该命令显示系统最近的告警日志，能发现端口震荡、协议中断等事件。

<Huawei> display trapbuffer
Info: Current total messages&nbsp;in&nbsp;log&nbsp;buffer is 10,&nbsp;in&nbsp;bitmap 10.
May 10 2023 14:25:32.123 Huawei %%01IFPDT/4/IF_DOWN(l)[0]:Interface GigabitEthernet0/0/1 has turned into DOWN state.
May 10 2023 14:25:33.456 Huawei %%01IFPDT/4/IF_UP(l)[1]:Interface GigabitEthernet0/0/1 has turned into UP state.
May 10 2023 14:25:34.789 Huawei %%01IFPDT/4/IF_DOWN(l)[2]:Interface GigabitEthernet0/0/1 has turned into DOWN state.
...

📌&nbsp;关键解读：

• 接口&nbsp;G0/0/1&nbsp;在1秒内频繁UP/DOWN，是典型的物理环路或网线故障
• 持续震荡会触发STP、MAC表刷新，导致CPU飙升

✅ 解决方案：检查该端口连接的设备，是否接了HUB或形成环路。

第四步：查看MAC地址表变化

——display mac-address

MAC表频繁刷新也会占用CPU资源。

<Huawei> display mac-address
--------------------------------------------------------------------------------
MAC Address &nbsp; &nbsp;VLAN &nbsp; &nbsp; &nbsp; Protocol &nbsp; &nbsp; Port/Type &nbsp; &nbsp; &nbsp; Aging/Status &nbsp; &nbsp;&nbsp;
--------------------------------------------------------------------------------
5489-980a-1234 10 &nbsp; &nbsp; &nbsp; &nbsp; Dynamic &nbsp; &nbsp; &nbsp;GigabitEthernet0/0/1 &nbsp;Yes/No
5489-980b-5678 20 &nbsp; &nbsp; &nbsp; &nbsp; Dynamic &nbsp; &nbsp; &nbsp;GigabitEthernet0/0/2 &nbsp;Yes/No
...
Total items displayed = 2048

📌&nbsp;排查点：

• Total items 接近设备上限（如4K、8K）：可能遭遇MAC泛洪攻击
• 大量Dynamic条目：正常；若突然增多，可能有非法设备接入
• 同一MAC出现在多个端口：存在环路或欺骗

第五步：查看协议状态

——display stp brief&nbsp;和&nbsp;display ospf peer

根据第二步的进程占用，针对性检查协议状态。

检查STP状态：

<Huawei> display stp brief
&nbsp;MSTID &nbsp;Port &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;Role &nbsp;STP State &nbsp; &nbsp; Protection
&nbsp; &nbsp;0 &nbsp; &nbsp;GigabitEthernet0/0/1 &nbsp; &nbsp; &nbsp; &nbsp;DESI &nbsp;FORWARDING &nbsp; &nbsp; &nbsp;NONE
&nbsp; &nbsp;0 &nbsp; &nbsp;GigabitEthernet0/0/2 &nbsp; &nbsp; &nbsp; &nbsp;ROOT &nbsp;FORWARDING &nbsp; &nbsp; &nbsp;NONE
&nbsp; &nbsp;0 &nbsp; &nbsp;GigabitEthernet0/0/3 &nbsp; &nbsp; &nbsp; &nbsp;ALTE &nbsp;DISCARDING &nbsp; &nbsp; &nbsp;NONE

• 观察是否有端口在&nbsp;DISCARDING&nbsp;和&nbsp;FORWARDING&nbsp;间频繁切换

检查OSPF邻居：

<Huawei> display ospf peer
&nbsp; &nbsp; &nbsp; &nbsp; OSPF Process 1 with Router ID 10.1.1.1
&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; Neighbor Brief Information

&nbsp;Area: 0.0.0.0 &nbsp; &nbsp; &nbsp; &nbsp;&nbsp;
&nbsp;Router ID &nbsp; &nbsp; &nbsp; Address &nbsp; &nbsp; &nbsp; &nbsp; State &nbsp; &nbsp; Dead Time &nbsp;Interface
&nbsp;10.1.1.2 &nbsp; &nbsp; &nbsp; &nbsp;192.168.1.2 &nbsp; &nbsp; Full &nbsp; &nbsp; &nbsp; &nbsp;32s &nbsp; &nbsp; &nbsp;GE0/0/0
&nbsp;10.1.1.3 &nbsp; &nbsp; &nbsp; &nbsp;192.168.1.3 &nbsp; &nbsp; Init &nbsp; &nbsp; &nbsp; &nbsp;10s &nbsp; &nbsp; &nbsp;GE0/0/0

• State&nbsp;频繁在&nbsp;Init、2-Way、Full&nbsp;间变化 → 邻居震荡

总结

🔚&nbsp;日常建议：

• 建立基线：记录正常状态下CPU各进程占用
• 配置告警：通过SNMP或iMaster NCE监控CPU异常
• 预防为主：启用端口安全、合理规划STP根桥、控制广播域

原创：老杨丨11年资深网络工程师，更多网工提升干货，请关注公众号：网络工程师俱乐部