BGP re-advertise received-routes back to peer #10052

DavePrashant · 2021-11-12T08:24:45Z

DavePrashant
Nov 12, 2021

I am using FRR version 8.2-dev (more details below). When run all daemons in “Datacenter” profile, then BGP re-advertise all received-routes back to peer.

Daemons:

frr       5156     1  0 06:27 ?        00:00:06 /usr/lib/frr/zebra -d -A 127.0.0.1 -s 90000000 -F datacenter
frr       5163     1  0 06:27 ?        00:00:05 /usr/lib/frr/staticd -d -A 127.0.0.1 -F datacenter
frr       5171     1  0 06:27 ?        00:00:12 /usr/lib/frr/bgpd -d -A 127.0.0.1 -F datacenter

FRR Version : FRRouting (version 8.2-dev)
Operating System: Linux Kernel 5.4.27-yocto-standard
Branch : master
Commit : cbdf030

[X] Did you check if this is a duplicate issue?
[X] Did you test it on the latest FRRouting/frr master branch?


**Configure Command:**
./configure  --build=x86_64-linux
     --host=x86_64-ion-linux
    --target=x86_64-ion-linux
    --prefix=/usr
    --exec_prefix=/usr
    --bindir=/usr/bin
    --sbindir=/usr/sbin
    --libexecdir=/usr/libexec
    --datadir=/usr/share
    --sysconfdir=/etc
    --sharedstatedir=/com
    --localstatedir=/var
    --libdir=/usr/lib
    --includedir=/usr/include
    --oldincludedir=/usr/include
    --infodir=/usr/share/info
    --mandir=/usr/share/man
    --disable-silent-rules
    --disable-dependency-tracking
    --localstatedir=/var/run/frr
    --sbindir=/usr/lib/frr
    --sysconfdir=/etc/frr
    --enable-bgpd
    --enable-configfile-mask=0640
    --enable-cumulus
    --enable-pimd
    --enable-fpm
    --enable-exampledir=/usr/share/doc/frr/examples/
    --enable-group=frr
    --enable-isisd
    --enable-irdp
    --enable-logfile-mask=0640
    --enable-multipath=64
    --enable-ospfclient=yes
    --enable-ospfd
    --enable-rtadv
    --enable-user=frr
    --enable-vty-group=frrvty
    --enable-vtysh
    --enable-watchfrr
    --disable-doc   ap_cv_cc_pie=no
    --disable-static
    --disable-capabilities
    --disable-systemd

Current configuration:

!
frr version 8.2-dev
frr defaults datacenter
log syslog informational
service integrated-vtysh-config

router bgp 1001
bgp router-id 10.16.122.142
neighbor 10.2.50.1 remote-as 1000
neighbor 10.2.50.1 description "classic local wan-side peer 16358936468590029"
neighbor 10.2.50.1 advertisement-interval 1
neighbor 10.2.50.1 timers 30 90
neighbor 10.2.50.1 timers connect 120
!
address-family ipv4 unicast
 network 10.2.52.0/24
 neighbor 10.2.50.1 soft-reconfiguration inbound
 maximum-paths 1
exit-address-family
exit

Received-routes

sh ip bgp neighbors 10.2.50.1 received-routes
BGP table version is 34, local router ID is 10.16.122.142, vrf id 0
Default local pref 100, local AS 1001
Status codes:  s suppressed, d damped, h history, * valid, > best, = multipath,
               i internal, r RIB-failure, S Stale, R Removed
Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
Origin codes:  i - IGP, e - EGP, ? - incomplete
RPKI validation codes: V valid, I invalid, N Not found

   Network          Next Hop            Metric LocPrf Weight Path
*> 10.2.1.0/24      10.2.50.1                              0 1000 65113 65112 65111 ?
*> 10.2.2.0/24      10.2.50.1                              0 1000 65113 65112 ?
*> 10.2.3.0/24      10.2.50.1                              0 1000 65113 65112 65111 ?
*> 10.2.4.0/24      10.2.50.1                              0 1000 65113 65112 ?
*> 10.2.5.0/24      10.2.50.1                              0 1000 65113 65112 ?
*> 10.2.6.0/24      10.2.50.1                              0 1000 65113 ?
*> 10.2.7.0/24      10.2.50.1                1             0 1000 ?
*> 10.2.8.0/24      10.2.50.1                              0 1000 65113 ?
*> 10.2.9.0/24      10.2.50.1                              0 1000 65113 ?
*> 10.2.14.0/24     10.2.50.1                              0 1000 65113 65212 65211 ?
*> 10.2.15.0/24     10.2.50.1                              0 1000 65113 65112 65111 i
*> 10.2.19.0/24     10.2.50.1                              0 1000 65113 65212 65211 ?
*> 10.2.20.0/24     10.2.50.1                              0 1000 65113 65212 ?
*> 10.2.21.0/24     10.2.50.1                              0 1000 65113 65212 65211 ?
*> 10.2.22.0/24     10.2.50.1                              0 1000 65113 65212 ?
*> 10.2.23.0/24     10.2.50.1                              0 1000 65113 65212 65211 ?
*> 10.2.24.0/24     10.2.50.1                              0 1000 65113 65212 ?
*> 10.2.25.0/24     10.2.50.1                              0 1000 65113 65212 65211 i
*> 10.2.49.0/24     10.2.50.1                1             0 1000 ?
*> 10.2.50.0/24     10.2.50.1                1             0 1000 ?
*> 10.2.51.0/24     10.2.50.1                1             0 1000 ?
*> 10.2.53.0/24     10.2.50.1                1             0 1000 ?
*> 10.2.54.0/24     10.2.50.1                1             0 1000 ?
*> 10.2.64.0/24     10.2.50.1                              0 1000 65113 2000 ?
*> 10.2.65.0/24     10.2.50.1                              0 1000 65113 2000 ?
*> 10.2.66.0/24     10.2.50.1                              0 1000 65113 2000 ?
*> 10.2.67.0/24     10.2.50.1                              0 1000 65113 2000 ?
*> 10.2.68.0/24     10.2.50.1                              0 1000 65113 2000 2001 i
*> 10.2.69.0/24     10.2.50.1                              0 1000 65113 2000 2001 i
*> 10.2.70.0/24     10.2.50.1                              0 1000 65113 2000 2001 i
*> 10.2.72.0/24     10.2.50.1                              0 1000 65113 2000 2001 i
*> 10.16.122.0/24   10.2.50.1                1             0 1000 ?

Total number of prefixes 32

advertised-routes
As per “network 10.2.52.0/24” only this prefix should get advertise, but all received-routes are also getting advertise.

sh ip bgp neighbors 10.2.50.1 advertised-routes
BGP table version is 34, local router ID is 10.16.122.142, vrf id 0
Default local pref 100, local AS 1001
Status codes:  s suppressed, d damped, h history, * valid, > best, = multipath,
               i internal, r RIB-failure, S Stale, R Removed
Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
Origin codes:  i - IGP, e - EGP, ? - incomplete
RPKI validation codes: V valid, I invalid, N Not found

   Network          Next Hop            Metric LocPrf Weight Path
*> 10.2.1.0/24      0.0.0.0                                0 1000 65113 65112 65111 ?
*> 10.2.2.0/24      0.0.0.0                                0 1000 65113 65112 ?
*> 10.2.3.0/24      0.0.0.0                                0 1000 65113 65112 65111 ?
*> 10.2.4.0/24      0.0.0.0                                0 1000 65113 65112 ?
*> 10.2.5.0/24      0.0.0.0                                0 1000 65113 65112 ?
*> 10.2.6.0/24      0.0.0.0                                0 1000 65113 ?
*> 10.2.7.0/24      0.0.0.0                                0 1000 ?
*> 10.2.8.0/24      0.0.0.0                                0 1000 65113 ?
*> 10.2.9.0/24      0.0.0.0                                0 1000 65113 ?
*> 10.2.14.0/24     0.0.0.0                                0 1000 65113 65212 65211 ?
*> 10.2.15.0/24     0.0.0.0                                0 1000 65113 65112 65111 i
*> 10.2.19.0/24     0.0.0.0                                0 1000 65113 65212 65211 ?
*> 10.2.20.0/24     0.0.0.0                                0 1000 65113 65212 ?
*> 10.2.21.0/24     0.0.0.0                                0 1000 65113 65212 65211 ?
*> 10.2.22.0/24     0.0.0.0                                0 1000 65113 65212 ?
*> 10.2.23.0/24     0.0.0.0                                0 1000 65113 65212 65211 ?
*> 10.2.24.0/24     0.0.0.0                                0 1000 65113 65212 ?
*> 10.2.25.0/24     0.0.0.0                                0 1000 65113 65212 65211 i
*> 10.2.49.0/24     0.0.0.0                                0 1000 ?
*> 10.2.50.0/24     0.0.0.0                                0 1000 ?
*> 10.2.51.0/24     0.0.0.0                                0 1000 ?
*> 10.2.52.0/24     0.0.0.0                  0         32768 i 
*> 10.2.53.0/24     0.0.0.0                                0 1000 ?
*> 10.2.54.0/24     0.0.0.0                                0 1000 ?
*> 10.2.64.0/24     0.0.0.0                                0 1000 65113 2000 ?
*> 10.2.65.0/24     0.0.0.0                                0 1000 65113 2000 ?
*> 10.2.66.0/24     0.0.0.0                                0 1000 65113 2000 ?
*> 10.2.67.0/24     0.0.0.0                                0 1000 65113 2000 ?
*> 10.2.68.0/24     0.0.0.0                                0 1000 65113 2000 2001 i
*> 10.2.69.0/24     0.0.0.0                                0 1000 65113 2000 2001 i
*> 10.2.70.0/24     0.0.0.0                                0 1000 65113 2000 2001 i
*> 10.2.72.0/24     0.0.0.0                                0 1000 65113 2000 2001 i
*> 10.16.122.0/24   0.0.0.0                                0 1000 ?

Total number of prefixes 33
28fcf610-8c5a-4be0-9f46-249ac78f92f8

Expected behavior
BGP should not re-advertise all received-routes back to peer.

Screenshots
None

Versions
Linux Kernel 5.4.27-yocto-standard

IgorAlov · 2021-11-12T19:13:09Z

IgorAlov
Nov 12, 2021

I don't think eBGP has any split horizon mechanism preventing this kind of route reflection ( not like iBGP--don't advertise iBGP learned routes to iBGP peers). So it seems like this behavior is legitimate.

I don't think it is a problem BGP needs to solve It is not really that big of a deal because routers should drop those routes due to AS_PATH loop protection.

0 replies

DavePrashant · 2021-11-12T19:29:45Z

DavePrashant
Nov 12, 2021
Author

Thanks IgorAlov for a quick response. Yes, I agree that eBGP doesn't have split horizon mechanism, however it would not advertise the routes learnt from the same peer unless it has better metric. In above case, I see those routes on my another peer with one AS-Num (i.e. 1001 1000 ... )

vyos@vyos:~$ sh ip bgp neighbors 10.2.50.2 received-routes
BGP table version is 0, local router ID is 10.2.7.2
Status codes: s suppressed, d damped, h history, * valid, > best, i - internal,
              r RIB-failure, S Stale, R Removed
Origin codes: i - IGP, e - EGP, ? - incomplete

   Network          Next Hop            Metric LocPrf Weight Path
*> 10.2.1.0/24      10.2.50.2                              0 1001 1000 65113 65112 65111 ?
*> 10.2.2.0/24      10.2.50.2                              0 1001 1000 65113 65112 ?
*> 10.2.3.0/24      10.2.50.2                              0 1001 1000 65113 65112 65111 ?
*> 10.2.4.0/24      10.2.50.2                              0 1001 1000 65113 65112 ?
*> 10.2.5.0/24      10.2.50.2                              0 1001 1000 65113 65112 ?
*> 10.2.6.0/24      10.2.50.2                              0 1001 1000 65113 ?
*> 10.2.7.0/24      10.2.50.2                              0 1001 1000 ?
*> 10.2.8.0/24      10.2.50.2                              0 1001 1000 65113 ?
*> 10.2.9.0/24      10.2.50.2                              0 1001 1000 65113 ?
*> 10.2.14.0/24     10.2.50.2                              0 1001 1000 65113 65212 65211 ?
*> 10.2.15.0/24     10.2.50.2                              0 1001 1000 65113 65112 65111 i
*> 10.2.19.0/24     10.2.50.2                              0 1001 1000 65113 65212 65211 ?
*> 10.2.20.0/24     10.2.50.2                              0 1001 1000 65113 65212 ?
*> 10.2.21.0/24     10.2.50.2                              0 1001 1000 65113 65212 65211 ?
*> 10.2.22.0/24     10.2.50.2                              0 1001 1000 65113 65212 ?
*> 10.2.23.0/24     10.2.50.2                              0 1001 1000 65113 65212 65211 ?
*> 10.2.24.0/24     10.2.50.2                              0 1001 1000 65113 65212 ?
*> 10.2.25.0/24     10.2.50.2                              0 1001 1000 65113 65212 65211 i
*> 10.2.49.0/24     10.2.50.2                              0 1001 1000 ?
*> 10.2.50.0/24     10.2.50.2                              0 1001 1000 ?
*> 10.2.51.0/24     10.2.50.2                              0 1001 1000 ?
*> 10.2.52.0/24     10.2.50.2                0             0 1001 i
*> 10.2.53.0/24     10.2.50.2                              0 1001 1000 ?
*> 10.2.54.0/24     10.2.50.2                              0 1001 1000 ?
*> 10.2.64.0/24     10.2.50.2                              0 1001 1000 65113 2000 ?
*> 10.2.65.0/24     10.2.50.2                              0 1001 1000 65113 2000 ?
*> 10.2.66.0/24     10.2.50.2                              0 1001 1000 65113 2000 ?
*> 10.2.67.0/24     10.2.50.2                              0 1001 1000 65113 2000 ?
*> 10.2.68.0/24     10.2.50.2                              0 1001 1000 65113 2000 2001 i
*> 10.2.69.0/24     10.2.50.2                              0 1001 1000 65113 2000 2001 i
*> 10.2.70.0/24     10.2.50.2                              0 1001 1000 65113 2000 2001 i
*> 10.2.72.0/24     10.2.50.2                              0 1001 1000 65113 2000 2001 i
*> 10.16.122.0/24   10.2.50.2                              0 1001 1000 ?

Total number of prefixes 33
vyos@vyos:~$

I do not see this behavior with other routing-stack we are using i.e vyatta VYOS 1.1.8 as well as Quagga BGPD 0.99.22. I agree that this Quagga version is pretty Old, but I assume that later should also have same behavior i.e. won't advertise the routes-learnt from same peer (to same peer) unless it has better route.

Another issue I notice was - if I change "profile" to "traditional" then FRR doesn't even advertise any route, even though "network" statements are present. I tried by enabling / disabling - "bgp network import-check". Hence, I changed my profiles to "Datacenter".

0 replies

IgorAlov · 2021-11-12T20:52:16Z

IgorAlov
Nov 12, 2021

Another issue I notice was - if I change "profile" to "traditional" then FRR doesn't even advertise any route, even though "network" statements are present. I tried by enabling / disabling - "bgp network import-check". Hence, I changed my profiles to "Datacenter".

It is work as it should. according to RFC8212 compliance without the incoming/outgoing filters, no routes will be accepted/advertised. You should use filters or you may use command no bgp ebgp-requires-policy to deactivate policy.

you may also look at:
https://docs.frrouting.org/en/latest/bgp.html#clicmd-neighbor-PEER-sender-as-path-loop-detection

0 replies

DavePrashant · 2021-11-13T00:54:57Z

DavePrashant
Nov 13, 2021
Author

I tried "neighbor <x.x.x.x> sender-as-path-loop-detection", and that didn't helped.
However, while going through FRR documentation, I found another option i.e. "neighbor <x.x.x.x> solo", and this seems to avoid advertising same prefixes back to the peer.

neighbor PEER solo

This command is used to indicate that routes advertised by the peer should not be reflected back to the peer. This command only is only meaningful when there is a single peer defined in the peer-group.

Thanks IgnorAlov. I am closing this issue.

2 replies

telirebarry Feb 16, 2022

Hi @DavePrashant,
I had the same problem as you and actually "neighbor <x.x.x.x> solo" solves the problem halfway..
I mean, after i put "neighbor <x.x.x.x> solo", and do "show run" this configuration does not appear.
So when i restart de frr service, the main problem reappears, I have to redo "neighbor <x.x.x.x> solo" to make it work again.

Do you remember having the same problem?

Thank you in advance for your answers

DavePrashant Mar 1, 2022
Author

No @telirebarry , I am not facing any such issue. Basically, in my environment, every time FRR (BGPD) restarts, we reconfigure everything. Hence, we may not be facing such issue.

donaldsharp · 2021-11-13T12:59:40Z

donaldsharp
Nov 13, 2021
Maintainer

When FRR received the initial addpath and update groups work. The decision was made to allow the retransmit of the +1 aspath routes back out. This allowed for much more efficient update group packing and much less work on FRR's side.

1 reply

wshamroukh Mar 1, 2022

I have similar setup where connected to two bgp peers:

!
router bgp 65016
 bgp router-id 172.16.0.4
 no bgp ebgp-requires-policy
 no bgp network import-check
 neighbor 10.1.0.4 remote-as 65515
 neighbor 10.1.0.4 description azgw1
 neighbor 10.1.0.4 ebgp-multihop 255
 neighbor 10.1.0.5 remote-as 65515
 neighbor 10.1.0.5 description azgw2
 neighbor 10.1.0.5 ebgp-multihop 255
 !
 address-family ipv4 unicast
  network 172.16.0.0/16
  neighbor 10.1.0.4 soft-reconfiguration inbound
  neighbor 10.1.0.5 soft-reconfiguration inbound
 exit-address-family
exit
!

here is the received-routes:

onprem-gw# sho ip bgp neighbors 10.1.0.4 received-routes
BGP table version is 3, local router ID is 172.16.0.4, vrf id 0
Default local pref 100, local AS 65016
Status codes:  s suppressed, d damped, h history, * valid, > best, = multipath,
               i internal, r RIB-failure, S Stale, R Removed
Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
Origin codes:  i - IGP, e - EGP, ? - incomplete
RPKI validation codes: V valid, I invalid, N Not found

   Network          Next Hop            Metric LocPrf Weight Path
*> 10.1.0.0/16      10.1.0.4                               0 65515 i

Total number of prefixes 1

i see the same received-routes are re-advertised:

onprem-gw# sho ip bgp neighbors 10.1.0.4 advertised-routes
BGP table version is 3, local router ID is 172.16.0.4, vrf id 0
Default local pref 100, local AS 65016
Status codes:  s suppressed, d damped, h history, * valid, > best, = multipath,
               i internal, r RIB-failure, S Stale, R Removed
Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
Origin codes:  i - IGP, e - EGP, ? - incomplete
RPKI validation codes: V valid, I invalid, N Not found

   Network          Next Hop            Metric LocPrf Weight Path
*> 10.1.0.0/16      0.0.0.0                                0 65515 i
*> 172.16.0.0/16    0.0.0.0                  0         32768 i

Total number of prefixes 2

I would expect only the network that i advertise (i.e network 172.16.0.0/16) to be only advertised

Is there a way to make sure not to re-advertise to the peer the received-routes that we received them from this peer?

donaldsharp · 2022-03-01T19:13:24Z

donaldsharp
Mar 1, 2022
Maintainer

http://docs.frrouting.org/en/latest/bgp.html#clicmd-neighbor-PEER-solo

2 replies

wshamroukh Mar 1, 2022

This works only if i have one peer as per the documentation. In my case I have two peers. When I have two peers, and i apply the neighbor <peer> solo we advertise the network that we wan to advertise (which what i am looking for), while we will re-advertise the received routes to the the second peer.. Here is a snippet of this configuration:

router bgp 65001
 bgp router-id 10.1.3.4
 no bgp ebgp-requires-policy
 neighbor 10.1.2.4 remote-as 65515
 neighbor 10.1.2.4 description rs2
 neighbor 10.1.2.4 solo
 neighbor 10.1.2.5 remote-as 65515
 neighbor 10.1.2.5 description rs1
 neighbor 10.1.2.5 solo
 !
 address-family ipv4 unicast
  network 0.0.0.0/1
  network 7.7.7.7/32
  network 128.0.0.0/1
  neighbor 10.1.2.4 soft-reconfiguration inbound
  neighbor 10.1.2.5 soft-reconfiguration inbound
 exit-address-family
exit

we advertise to this peer the networks that we want to advertise - which what I want:

nva# sho ip bgp neighbors 10.1.2.5 advertised-routes
BGP table version is 9, local router ID is 10.1.3.4, vrf id 0
Default local pref 100, local AS 65001
Status codes:  s suppressed, d damped, h history, * valid, > best, = multipath,
               i internal, r RIB-failure, S Stale, R Removed
Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
Origin codes:  i - IGP, e - EGP, ? - incomplete
RPKI validation codes: V valid, I invalid, N Not found

   Network          Next Hop            Metric LocPrf Weight Path
*> 0.0.0.0/1        0.0.0.0                  0         32768 i
*> 7.7.7.7/32       0.0.0.0                  0         32768 i
*> 128.0.0.0/1      0.0.0.0                  0         32768 i

Total number of prefixes 3

While the second peer, we advertise what we receive as well:

nva# sho ip bgp neighbors 10.1.2.4 advertised-routes
BGP table version is 9, local router ID is 10.1.3.4, vrf id 0
Default local pref 100, local AS 65001
Status codes:  s suppressed, d damped, h history, * valid, > best, = multipath,
               i internal, r RIB-failure, S Stale, R Removed
Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
Origin codes:  i - IGP, e - EGP, ? - incomplete
RPKI validation codes: V valid, I invalid, N Not found

   Network          Next Hop            Metric LocPrf Weight Path
*> 0.0.0.0/1        0.0.0.0                  0         32768 i
*> 7.7.7.7/32       0.0.0.0                  0         32768 i
*> 10.1.0.0/16      0.0.0.0                                0 65515 i
*> 10.11.1.0/24     0.0.0.0                                0 65515 i
*> 10.12.1.0/24     0.0.0.0                                0 65515 i
*> 128.0.0.0/1      0.0.0.0                  0         32768 i

Total number of prefixes 6

sdntechforum Apr 18, 2023

I have the same problem with ""frr defautt traditional" profile

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

BGP re-advertise received-routes back to peer #10052

{{title}}

Replies: 6 comments 5 replies

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

Select a reply

BGP re-advertise received-routes back to peer #10052

DavePrashant Nov 12, 2021

Replies: 6 comments · 5 replies

IgorAlov Nov 12, 2021

DavePrashant Nov 12, 2021 Author

IgorAlov Nov 12, 2021

DavePrashant Nov 13, 2021 Author

telirebarry Feb 16, 2022

DavePrashant Mar 1, 2022 Author

donaldsharp Nov 13, 2021 Maintainer

wshamroukh Mar 1, 2022

donaldsharp Mar 1, 2022 Maintainer

wshamroukh Mar 1, 2022

sdntechforum Apr 18, 2023

DavePrashant
Nov 12, 2021

Replies: 6 comments 5 replies

IgorAlov
Nov 12, 2021

DavePrashant
Nov 12, 2021
Author

IgorAlov
Nov 12, 2021

DavePrashant
Nov 13, 2021
Author

DavePrashant Mar 1, 2022
Author

donaldsharp
Nov 13, 2021
Maintainer

donaldsharp
Mar 1, 2022
Maintainer