I will send you the detailed topology and explain what is the expected behavior, that is not occuring.
Em dom, 2 de set de 2018 04:53, Maria Jan Matějka <[email protected]> escreveu: > By default yes. I don't know why your peers do announce the actual routes > and what is written in your config and what you want. What you provided, > seems legit by default. > > Maria > > On September 2, 2018 8:59:17 AM GMT+02:00, Marcio < > [email protected]> wrote: >> >> Yes, i have. This is a problem? BGP don't avoid loops automatically >> finding it own ASN in AS PATH? >> >> Em dom, 2 de set de 2018 03:27, Maria Jan Matějka <[email protected]> >> escreveu: >> >>> >>> >>> Now there is one route originated from R. >>> >>> >> > 2018-09-01 21:21:08.000 <TRACE> SDNRTR: Got UPDATE >>> >> > 2018-09-01 21:21:08.000 <TRACE> SDNRTR > added [best] 10.3.1.0/24 >>> >> > unicast >>> >> > 2018-09-01 21:21:08.000 <TRACE> kernel1 < replaced 10.3.1.0/24 >>> >unicast >>> >> > 2018-09-01 21:21:08.000 <TRACE> SDNRTR < rejected by protocol >>> >> > 10.3.1.0/24 >>> >> > unicast >>> >> > 2018-09-01 21:21:08.000 <TRACE> R8 < replaced 10.3.1.0/24 unicast >>> >> > 2018-09-01 21:21:08.000 <TRACE> R8: Sending UPDATE >>> >>> Here S added better route and it was propagated to kernel and R. S >>> doesn't send it back. >>> >>> >> > 2018-09-01 21:21:08.000 <TRACE> SDNRTR: Got UPDATE >>> >> > 2018-09-01 21:21:08.000 <TRACE> SDNRTR > removed [replaced] >>> >> > 10.3.1.0/24 >>> >> > unicast >>> >> > 2018-09-01 21:21:08.000 <TRACE> kernel1 < replaced 10.3.1.0/24 >>> >unicast >>> >> > 2018-09-01 21:21:08.000 <TRACE> SDNRTR < added 10.3.1.0/24 unicast >>> >> > 2018-09-01 21:21:08.000 <TRACE> R8 < rejected by protocol >>> >10.3.1.0/24 >>> >> > unicast >>> >> > 2018-09-01 21:21:08.000 <TRACE> R8 < removed 10.3.1.0/24 unicast >>> >> > 2018-09-01 21:21:08.000 <TRACE> SDNRTR: Sending UPDATE >>> >> > 2018-09-01 21:21:08.000 <TRACE> R8: Sending UPDATE >>> >>> Here S sent another route (replacing its previous one) which came out >>> worse even than the previous from R. So the R's route is announced to S and >>> withdraw to R. >>> >>> >> > 2018-09-01 21:21:08.001 <TRACE> R8: Got UPDATE >>> >> > 2018-09-01 21:21:08.001 <TRACE> R8 > removed [sole] 10.3.1.0/24 >>> >> > unicast >>> >> > 2018-09-01 21:21:08.001 <TRACE> kernel1 < removed 10.3.1.0/24 >>> >unicast >>> >> > 2018-09-01 21:21:08.001 <TRACE> SDNRTR < removed 10.3.1.0/24 >>> >unicast >>> >> > 2018-09-01 21:21:08.001 <TRACE> SDNRTR: Sending UPDATE >>> >>> R8 withdraws its route. >>> >>> >> > 2018-09-01 21:21:08.001 <TRACE> R8: Got UPDATE >>> >> > 2018-09-01 21:21:08.001 <TRACE> R8 > added [best] 10.3.1.0/24 >>> >unicast >>> >> > 2018-09-01 21:21:08.001 <TRACE> kernel1 < added 10.3.1.0/24 unicast >>> >> > 2018-09-01 21:21:08.001 <TRACE> SDNRTR < added 10.3.1.0/24 unicast >>> >> > 2018-09-01 21:21:08.001 <TRACE> R8 < rejected by protocol >>> >10.3.1.0/24 >>> >>> R8 sends a route and BIRD sends update to S. >>> >>> Everything in the log is OK, just the peers (R and S) are doing quite >>> strange things. Don't you have loops in your routing? >>> >>> Maria >>> >> > -- > Sent from my Android device with K-9 Mail. Please excuse my brevity. >
