MPLS FIB differente from OSPF FIB

RouterOS Forwarding Protocols
1

Hi,

I noticed a host inside network 172.16.16.0/24 (172.16.16.254) wasn’t replying to Dude Server.

After do a traceroute from Dude Server to 172.16.16.254, a loop was detected:

tool traceroute address=172.16.16.254
 # ADDRESS                          LOSS SENT    LAST     AVG    BEST   WORST STD-DEV STATUS             
 1 179.125.47.225                     0%    1   1.3ms     1.3     1.3     1.3       0                    
 2 172.16.0.89                        0%    1     1ms       1       1       1       0                    
 3 172.16.200.1                       0%    1   2.2ms     2.2     2.2     2.2       0 <MPLS:L=16823,E=0> 
 4 172.16.0.9                         0%    1   2.2ms     2.2     2.2     2.2       0 <MPLS:L=2080,E=0>  
 5 172.16.0.29                        0%    1   2.4ms     2.4     2.4     2.4       0                    
 6 172.16.0.9                         0%    1   2.3ms     2.3     2.3     2.3       0 <MPLS:L=2080,E=0>  
 7 172.16.0.29                        0%    1   2.4ms     2.4     2.4     2.4       0                    
 8 172.16.0.9                         0%    1   2.6ms     2.6     2.6     2.6       0 <MPLS:L=2080,E=0>  
 9 172.16.0.29                        0%    1   2.4ms     2.4     2.4     2.4       0                    
10 172.16.0.9                         0%    1   2.4ms     2.4     2.4     2.4       0 <MPLS:L=2080,E=0>  
11 172.16.0.29                        0%    1   2.4ms     2.4     2.4     2.4       0                    
12 172.16.0.9                         0%    1   2.8ms     2.8     2.8     2.8       0 <MPLS:L=2080,E=0>  
13 172.16.0.29                        0%    1   3.2ms     3.2     3.2     3.2       0                    
14 172.16.0.9                         0%    1   2.2ms     2.2     2.2     2.2       0 <MPLS:L=2080,E=0>  
15 172.16.0.29                        0%    1   2.7ms     2.7     2.7     2.7       0                    
16 172.16.0.9                         0%    1   2.8ms     2.8     2.8     2.8       0 <MPLS:L=2080,E=0>  
17 172.16.0.29                        0%    1   2.3ms     2.3     2.3     2.3       0

Then I investigated 172.16.0.9 (where loop starts) to try understand what was occurring:

Then I noticed OSPF was pointing the 172.16.16.0/24 network to the correct gateway, but MPLS was pointing to another interface, generating the loop.

Is there a way to remove that line from LFIB? I’ve tried to remove, but it’s dynamic and do not allowed to be removed

:confused:

2

We have the same problem. Use devices ccr 1036 and rb3011 in mpls cloud (ldp). After restarting ccr1036 problem going out, but after same time it came back.
we try to use many versions of ROS in both devices, 6.41.3-6.43.2 but it’s dont take effect

3

Could you provide feedback after matching or reducing MPLS LDP interface hello and dead timers to be smaller or equal to OSPF timers?

I’ve not had a re-occurance since matching my timers to that of the OSPF interfaces and assigning labels out of reserved ranges.

More detail here:

http://forum.mikrotik.com/t/mpls-incorrect-forwarding-table/103749/15