MSR56 OSPF邻居停留在在init状态较长时间

组网：MSR5660 (2/3/2)----8M MSTP专线（过传输）---- (3/2/6)SR6604

两端ospf邻居报DeadInterval timer expired（这个是传输问题，两端设备的入接口下都有CRC，需联系运营商排查）。此时，SR66侧OSPF直接由fulll 变为down，LLDP也是马上超时。但MSR56侧OSPF状态由FULL 变为 INIT状态，且持续了2小时左右，并且LLDP也是2小时左右后才报超时日志的。 

SR66：

%Jan 27 20:06:31:165 2022 DHJ-WAN-RT-6604-02 OSPF/6/OSPF_LAST_NBR_DOWN: OSPF 10255 Last neighbor down event: Router ID: 10.216.0.2 Local address: 10.248.250.217 Remote address: 10.248.250.218 Reason: DeadInterval timer expired.

%Jan 27 20:06:31:165 2022 DHJ-WAN-RT-6604-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.218(GigabitEthernet3/2/6) changed from FULL to DOWN.

%Jan 27 20:07:39:276 2022 DHJ-WAN-RT-6604-02 LLDP/5/LLDP_NEIGHBOR_AGE_OUT: -Slot=3; Nearest bridge agent neighbor aged out on port GigabitEthernet3/2/6 (IfIndex 1095), neighbor's chassis ID is 9429-2fd3-82b0, port ID is GigabitEthernet2/3/2.

MSR56：

%Jan 27 20:06:34:167 2022 KF-RT-MSR5660-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.217(GigabitEthernet2/3/2) changed from FULL to INIT.

%Jan 27 22:49:54:549 2022 KF-RT-MSR5660-02 OSPF/6/OSPF_LAST_NBR_DOWN: OSPF 10255 Last neighbor down event: Router ID: 10.248.255.3 Local address: 10.248.250.218 Remote address: 10.248.250.217 Reason: DeadInterval timer expired.

%Jan 27 22:49:54:549 2022 KF-RT-MSR5660-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.217(GigabitEthernet2/3/2) changed from INIT to DOWN.

#  

ospf 10255 router-id 10.216.0.2

area 0.0.0.216

  network 10.216.0.2 0.0.0.0

  network 10.216.0.10 0.0.0.0

  network 10.216.0.25 0.0.0.0

  network 10.248.250.218 0.0.0.0

  network 10.248.250.222 0.0.0.0

  network 10.248.250.226 0.0.0.0

#

interface GigabitEthernet2/3/2

port link-mode route

bandwidth 8000

ip address 10.248.250.218 255.255.255.252

ospf network-type p2p

packet-filter 3011 outbound

#

1. 故障现象发生在27日20:06:34，MSR56侧OSPF邻居full->init，SR66侧OSPF邻居full->down。
2. SR66邻居由full至down，是因为超过4个hello间隔没有收到来自MSR56的hello，保活超时，判断依据为下面日志。

%Jan 27 20:06:31:165 2022 DHJ-WAN-RT-6604-02 OSPF/6/OSPF_LAST_NBR_DOWN: OSPF 10255 Last neighbor down event: Router ID: 10.216.0.2 Local address: 10.248.250.217 Remote address: 10.248.250.218 Reason: DeadInterval timer expired.

%Jan 27 20:06:31:165 2022 DHJ-WAN-RT-6604-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.218(GigabitEthernet3/2/6) changed from FULL to DOWN.

3. 之后SR66会继续发hello，hello中本地router id正常携带，neighbor id为0.0.0.0。MSR56收到该hello后，看到其中neighbor id非MSR56 router id，因此MSR56邻居full->init。

%Jan 27 20:06:34:167 2022 KF-RT-MSR5660-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.217(GigabitEthernet2/3/2) changed from FULL to INIT.

4. 从上述现象分析，故障时间点MSR56至SR66方向不通，SR66至MSR56方向通。从线路异常至SR66侧OSPF down，时间为30-40s之间。

MSR56：

1. 至22:49:54，MSR56侧OSPF邻居init->down，说明此时SR66至MSR56方向也不通了，MSR56收不到hello持续4个hello间隔后，邻居down。

%Jan 27 22:49:54:549 2022 KF-RT-MSR5660-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.217(GigabitEthernet2/3/2) changed from INIT to DOWN.

2. 之后MSR56侧有数次init->down，说明期间SR66至MSR56方向时通时断，通时MSR56邻居init，断时邻居down。

%Jan 27 22:52:44:549 2022 KF-RT-MSR5660-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.217(GigabitEthernet2/3/2) changed from INIT to DOWN.

%Jan 27 22:53:54:549 2022 KF-RT-MSR5660-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.217(GigabitEthernet2/3/2) changed from INIT to DOWN.

%Jan 27 23:02:44:549 2022 KF-RT-MSR5660-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.217(GigabitEthernet2/3/2) changed from INIT to DOWN.

3. 至23:35:44，MSR56侧OSPF邻居full。

%Jan 27 23:35:44:244 2022 KF-RT-MSR5660-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.217(GigabitEthernet2/3/2) changed from LOADING to FULL.

SR66：

1. 对应SR66侧日志，23:33:21邻居init->down，说明此前MSR56到SR66方向通过，SR66收到了来自MSR56的hello，但neighbor为空，因此邻居init。23:33:21，MSR56至SR66方向再次不通，SR66邻居init->down。

%Jan 27 23:33:21:165 2022 DHJ-WAN-RT-6604-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.218(GigabitEthernet3/2/6) changed from INIT to DOWN.

%Jan 27 23:35:11:165 2022 DHJ-WAN-RT-6604-02 OSPF/6/OSPF_LAST_NBR_DOWN: OSPF 10255 Last neighbor down event: Router ID: 10.216.0.2 Local address: 10.248.250.217 Remote address: 10.248.250.218 Reason: DeadInterval timer expired.

%Jan 27 23:35:11:165 2022 DHJ-WAN-RT-6604-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.218(GigabitEthernet3/2/6) changed from INIT to DOWN.

2. 至23:35:44，两个方向都恢复互通，邻居full。

%Jan 27 23:35:44:246 2022 DHJ-WAN-RT-6604-02 OSPF/5/OSPF_NBR_CHG: OSPF 10255 Neighbor 10.248.250.218(GigabitEthernet3/2/6) changed from LOADING to FULL.

3. 从线路恢复至OSPF  p2p邻居full，需要数秒时间。OSPF路由收敛速度取决于路由量，现网设备路由数量1k以下，收敛时间也在秒级。

查看两端设备诊断，两端接口均没有发生过振荡，不会停止发送OSPF hello。SR66接口G3/2/6入方向有大量错包，可能与此次故障有关，如MSR56发出报文到达SR66时CRC校验错误，SR66就不会上送CPU，表现为hello超时。从诊断中无法判断错包产生时间，需要在故障期间dis interface g3/2/6查看入方向错包是否持续增长。

GigabitEthernet2/3/2

Last link flapping: 76 weeks 3 days 0 hours 5 minutes

Input (total):  7661561159 packets, 1108441248915 bytes

         - unicasts, - broadcasts, - multicasts, - pauses

Input (normal):  7661561155 packets, - bytes

         7626040521 unicasts, 46 broadcasts, 35520588 multicasts, 0 pauses

Input:  4 input errors, 0 runts, 0 giants, - throttles

         4 CRC, - frame, - overruns, 0 aborts

         - ignored, - parity errors

Output (total): 9118312095 packets, 1187922082804 bytes

         - unicasts, - broadcasts, - multicasts, - pauses

Output (normal): 9118312095 packets, - bytes

         9082291152 unicasts, 75 broadcasts, 36020868 multicasts, 0 pauses

Output: 0 output errors, - underruns, - buffer failures

         0 aborts, 0 deferred, 0 collisions, 0 late collisions

         - lost carrier, - no carrier

GigabitEthernet3/2/6

Last link flapping: 70 weeks 0 days 0 hours 56 minutes

Input (total):  8062889283 packets, 1059672504493 bytes

         8029029602 unicasts, 4 broadcasts, 31288937 multicasts, - pauses

Input (normal):  8060318543 packets, 1059078251094 bytes

         8029029602 unicasts, 4 broadcasts, 31288937 multicasts, - pauses

Input:  2570740 input errors, 0 runts, 0 giants, - throttles

         2570740 CRC, - frame, 0 overruns, - aborts

         - ignored, - parity errors

Output (total): 6975777337 packets, 1038323619478 bytes

         6944555507 unicasts, 24 broadcasts, 31221806 multicasts, - pauses

Output (normal): 6975777337 packets, 1038323619478 bytes

         6944555507 unicasts, 24 broadcasts, 31221806 multicasts, - pauses

Output: 0 output errors, 0 underruns, - buffer failures

         - aborts, 0 deferred, 0 collisions, 0 late collisions

         0 lost carrier, - no carrier

建议排查中间线路，主要排查MSR56至SR66方向。

后续可以考虑增加OSPF BFD加快感知线路异常，BFD探测间隔用默认的400ms * 5即可。配置BFD后，如果线路频繁振荡，可能导致OSPF频繁振荡。