MP Group成员接口状态频繁Up/Down的故障解决步骤如下：
网络环境
两台NE40路由器RouterA和RouterB通过MP Group链路连接，MP Group由8条E1链路捆绑而成。如图所示。
MP Group故障组网图
在进行一次设备调整后，发现业务不通，查看接口发现MP Group成员不停地Up/Down，MP Group自身也有Up/Down现象。
查看日志缓冲区，可以看到有很多Serial口和MP Group接口Up/Down的日志信息：
- %Jul 21 04:06:48 2007 A2 LSPM/5/LSPM:
- Info:[LSPM] Get interface event 1 and address 0x0 for interface Mp-group1/1/1
- %Jul 21 04:06:48 2007 A2 IFNET/5/UPDOWN:Interface Serial1/1/3:0 Turns into DOWN state
- %Jul 21 04:06:48 2007 A2 LSPM/5/LSPM:
- Info:[LSPM] Get interface event 1 and address 0x0 for interface Mp-group1/1/1
- %Jul 21 04:06:48 2007 A2 IFNET/5/UPDOWN:Interface Serial1/1/2:0 Turns into DOWN state
- %Jul 21 04:06:48 2007 A2 LSPM/5/LSPM:
- Info:[LSPM] Get interface event 1 and address 0x0 for interface Mp-group1/1/1
- %Jul 21 04:06:48 2007 A2 IFNET/5/UPDOWN:Interface Serial1/1/1:0 Turns into DOWN state
- %Jul 21 04:06:48 2007 A2 LSPM/5/LSPM:
- Info:[LSPM] Get interface event 1 and address 0x0 for interface Mp-group1/1/1
- %Jul 21 04:06:48 2007 A2 IFNET/5/UPDOWN:Interface Serial1/1/0:0 Turns into DOWN state
- %Jul 21 04:06:48 2007 A2 LSPM/5/LSPM:
- Info:[LSPM] Get interface event 1 and address 0x0 for interface Mp-group1/1/1
- %Jul 21 04:06:47 2007 A2 IFNET/5/UPDOWN:Line protocol(MPLSCP) on interface Mp-group1/1/1, changed state to UP
- %Jul 21 04:06:47 2007 A2 RM/4/RMLOG:OSPF PhyIf Info: IfName: Mp-group1/1/1, P1: 4d, P2: 400088e, P3: 3, P4: 10240000, P5: 1500, P6: 4
- %Jul 21 04:06:47 2007 A2 RM/4/RMLOG:OSPF LogicIf Info: IfName: Mp-group1/1/1, P1: 21, P2: 77, P3: 67111054, P4: 3, P5: 10240000, P6: 1500, P7: 4
- %Jul 21 04:06:47 2007 A2 RM/4/RMLOG:OSPF PhyIf Info: IfName: Mp-group1/1/1, P1: 4d, P2: 400088e, P3: 3, P4: 10240000, P5: 1500, P6: 6
- %Jul 21 04:06:47 2007 A2 IFNET/5/UPDOWN:Line protocol(OSICP) on interface Mp-group1/1/1, changed state to UP
- %Jul 21 04:06:47 2007 A2 IFNET/5/UPDOWN:Line protocol(MPLSCP) on interface Mp-group1/1/1, changed state to DOWN
故障分析
步骤 1 检查链路两端NE40的配置，发现配置没有异常。
步骤 2 由于设备调整后链路中断，设备重起后PPP需要重新协商，可能出现协商不成功，重启接口板后发现问题依然存在。
步骤 3 执行命令display interface mp-group，查看接口信息，发现两端路由器上的MP group的成员接口中有一些处在Active状态，另一些处在Inactive状态，并且状态不停的变化，而正常情况下应该都是Active。同时发现Peer IP address协商异常。显示的接口信息如下：
- <A2> display interface mp-group1/1/1
- Mp-group1/1/1 current state: up,
- Line protocol current state: up
- Description: to_B2_MPgroup1/0/1
- Internet address is 10.240.240.5/30
- Link-protocol is PPP, MP, NO SHORTSEQUENCE, DISCRIMINATOR
- MRRU 1500 bytes, Peer MRRU 1500 bytes
- The Max HQOS BandWidth is 9520 Kbits
- Include: 5 * E1, Active: 4 * E1
- Serial1/1/4:0 A
- Serial1/1/3:0 A
- Serial1/1/2:0 A
- Serial1/1/1:0 A
- Serial1/1/0:0 I
- Peer IP address is 10.240.240.2/32
- NCP: IPCP Opened, OSICP Opened, MPLSCP Opened
- <A1> display interface mp-group1/1/1
- Mp-group1/1/1 current state: up,
- Line protocol current state: up
- Description:to_B1_MPgroup1/0/1
- Internet address is 10.240.240.1/30
- Link-protocol is PPP, MP, NO SHORTSEQUENCE, DISCRIMINATOR
- MRRU 1500 bytes, Peer MRRU 1500 bytes
- The Max HQOS BandWidth is 9520 Kbits
- Include: 5 * E1, Active: 3 * E1
- Serial1/1/4:0 A
- Serial1/1/3:0 A
- Serial1/1/2:0 I
- Serial1/1/1:0 A
- Serial1/1/0:0 I
- Peer IP address is 10.240.240.6/32
- NCP: IPCP Opened, OSICP Opened, MPLSCP Opened
步骤 4 通过上面显示的信息可以判断，这是由于中间传输链路交叉连接，导致两个MP Group交叉。
----结束
处理步骤
步骤 1 检查传输线路，找到传输连接交叉的线路，调整传输线路。
步骤 2 观察路由器上MP Group链路的状态，保持在Up状态，故障得到排除。
----结束
案例总结
在处理MP Group链路的故障时，也要检查中间传输线路的故障。
【编辑推荐】
- 路由器故障：CE1接口网速较慢
- 路由器故障：POS口状态反复Up/Down
- 路由器故障：POS接口修改MTU值无法生效
- 路由器故障：POS接口无法收到对端的报文
- 路由器故障：POS链路不能Up：C2字节不匹配
