[j-nsp] BGP troubleshooting
Campbell, Alex
Alex.Campbell at ogilvy.com.au
Sat Nov 1 06:47:30 EDT 2008
Hi all,
We have run into a BGP issue bringing up a new ISP on our J4350s and I
have been struggling to figure out what's going on. If anyone has any
troubleshooting suggestions or ideas as to what might be doing wrong, I
would be most grateful.
We have two border routers, each connected to a separate port with the
same ISPs. The interface IPs on those ports are
border1: 1.1.1.1.76/29
border2: 1.1.1.1.77/29
border1 is configured to bring up a multihop BGP session with ISP's
router at 2.2.2.2. We have added a static route on border1 to
2.2.2.2/32 with a next-hop of 1.1.1.1.73. This works correctly and
border1 receives a full table of around 251,000 routes.
border2 is configured to bring up a multihop BGP session with ISP's
router at 3.3.3.3. We have added a static route on border2 to
3.3.3.3/32 with a next-hop of 1.1.1.1.73.
The problem is that when we allow border2 to bring up its BGP session
with the ISP router, the session comes up successfully but the device
receives 0 routes from the ISP router, and we lose complete internet
connectivity.
This is what we see on border2 after bringing up the session:
=============================================================
alex at border2# run show bgp summary
Groups: 2 Peers: 2 Down peers: 0
Table Tot Paths Act Paths Suppressed History Damp State
Pending
inet.0 250985 250972 0 0 0
0
Peer AS InPkt OutPkt OutQ Flaps Last Up/Dwn
State|#Active/Received/Damped...
5.5.5.1 24438 67593 103 0 0 43:56
250972/250985/0 0/0/0
3.3.3.3 9443 1 4 0 0 20 0/0/0
0/0/0
=============================================================
Here's an example traceroute of what happens from a machine on our
network after we bring up border2's BGP session to ISP:
=============================================================
C:\Users\a>tracert www.google.com.au
Tracing route to www.l.google.com [209.85.171.147]
over a maximum of 30 hops:
1 <1 ms <1 ms <1 ms red2.fw1.mel1.x.com.au [6.6.6.1]
2 23 ms 28 ms 29 ms border2 [5.5.5.35]
3 23 ms 29 ms 29 ms ge-0-0-2.border1.mel1.x.com.au
[5.5.5.17]
4 1 ms 1 ms 1 ms 075.196.idc.ourisp.net.au [1.1.1.1.75]
5 * * * Request timed out.
6 * * * Request timed out.
7 * * * Request timed out.
8 * * * Request timed out.
=============================================================
Here's an example of what happens from a machine on our network after we
shutdown border2's BGP session to ISP:
C:\Users\a>tracert www.google.com.au
Tracing route to www.l.google.com [209.85.171.103]
over a maximum of 30 hops:
1 <1 ms <1 ms <1 ms red2.fw1.mel1.x.com.au [6.6.6.1]
2 27 ms 29 ms 29 ms border2 [5.5.5.35]
3 25 ms 29 ms 18 ms ge-0-0-2.border1.mel1.x.com.au
[5.5.5.17]
4 1 ms 1 ms 1 ms 075.196.idc.ourisp.net.au [1.1.1.1.75]
5 1 ms 1 ms 1 ms ge4-15.sw02.mel.ourisp.net.au
[210.50.0.9]
6 132 ms 2 ms 2 ms 210.50.233.202
7 170 ms 170 ms 170 ms p3-0.pr1-lax.primustel.com
[209.227.149.241]
8 170 ms 173 ms 171 ms g9-0.pr1-lax.primustel.com
[209.227.129.141]
9 172 ms 172 ms 172 ms 209.227.129.62
10 172 ms 174 ms 173 ms core2-1-1-0.pao.net.google.com
[198.32.176.31]
11 175 ms 175 ms 174 ms 216.239.49.250
12 210 ms 198 ms 197 ms 216.239.47.186
13 205 ms 189 ms 195 ms 216.239.46.211
14 190 ms 192 ms 194 ms 64.233.174.97
15 191 ms 200 ms 190 ms 209.85.251.141
16 189 ms 190 ms 203 ms 74.125.31.134
17 189 ms 191 ms 190 ms cg-in-f103.google.com [209.85.171.103]
Trace complete.
=============================================================
More information about the juniper-nsp
mailing list