Routing Loop - problem

[Jon Are Endrerud]

Hey everyone.

I have a routing problem. Seems like a routing loop of some kind. We are in a prosess of changing topology and I had to be creative, but maybe to creative.

When a unit is turned on, it respons normaly to ping, but when is off you get TTL expired in transit. 10.16.160.0/22 is network that im trying to reach.

Tracert shows this:

 32.54.20.97
 32.54.20.2
 10.16.108.10
 10.16.108.2
 10.16.108.10
 10.16.108.2
 10.16.108.10
 10.16.108.2
 10.16.108.10
 10.16.108.2

If the a unit on the other end is turned on the tracert is normal.

Attached are a quick hand drawing. I understand there is a problem, but not how I can do this in another way

Thanx in advance

Jon

[Rick Morris]

If you traceroute to any IP on the 10.16.160.0/22 network, which on the drawing shows VIKEN it works when the device is on, as in the other end router from the interface 10.16.108.2/22?

What does the routing table show on both ends?

What it looks like based on the info above is that there is a route back toward 108.2 with the 10.16.160.0 network and when there is a route back toward 108.10, are you using static routes?

[Jon Are Endrerud]

Yes this is static routes.

The 10.16.160.0 network has a Procurve 2620 as router with ip 10.16.160.1, and a 0.0.0.0/0 route to 10.16.108.2 and a static ip in the same vlan 10.16.108.10.

Attached are the config on 10.16.108.2

And "sh ip route"

Cisco 2901:

Gateway of last resort is 10.16.108.1 to network 0.0.0.0
S*    0.0.0.0/0 [1/0] via 10.16.108.1
      10.0.0.0/8 is variably subnetted, 6 subnets, 3 masks
C        10.0.0.0/23 is directly connected, Vlan1
L        10.0.0.7/32 is directly connected, Vlan1
C        10.16.108.0/22 is directly connected, GigabitEthernet0/0
L        10.16.108.2/32 is directly connected, GigabitEthernet0/0
S        10.16.160.0/22 [1/0] via 10.16.108.10
S        10.16.198.0/23 [1/0] via 10.16.108.11
      32.0.0.0/8 is variably subnetted, 2 subnets, 2 masks
C        32.54.16.0/20 is directly connected, GigabitEthernet0/1
L        32.54.20.2/32 is directly connected, GigabitEthernet0/1
      155.55.0.0/24 is subnetted, 2 subnets
S        155.55.2.0 [1/0] via 32.54.20.97
S        155.55.10.0 [1/0] via 32.54.20.97
S     192.168.1.0/24 [1/0] via 32.54.20.97
S     192.168.4.0/24 [1/0] via 32.54.20.97
S     192.168.5.0/24 [1/0] via 32.54.20.97

Procurve:

Destination        Gateway         VLAN Type      Sub-Type   Metric     Dist.
  ------------------ --------------- ---- --------- ---------- ---------- -----
  0.0.0.0/0          10.16.108.2     5    static               1          1    
  10.0.0.0/23        admin           1    connected            1          0    
  10.16.108.0/22     klient          5    connected            1          0    
  10.16.160.0/22     int-viken       280  connected            1          0    
  127.0.0.0/8        reject               static               0          0    
  127.0.0.1/32       lo0                  connected            1          0    
 

What is strange is that i have the same setup against 10.16.198.0 (ref: 2901 config) but it show no sign of a loop.

Thanks for you reply.

[Rick Morris]

OK, so there is a static route on 108.2 for 10.16.160

When online everything is working fine when not online you get a response that looks like traffic is bouncing back and forth. So the 16.160.x network is not accessible so the default route is sending the traffic back, it appears that on the connected procurve that the network looks like it is no longer present. So instead of dropping the traffic it is sent back and then a route look-up takes place and sent back, now you have the loop.

What does a ping indicate, I would suspect you will get ..... instead of the !!!!! or u.u.u

[Jon Are Endrerud]

I agree.

The procuve debug says:

IP  mIpPktRecv:s=32.54.20.8(vl5) d=10.16.160.3 drop ttl exceeded.

Ping from 32.54.20.x says:

Pinging 10.16.160.3 with 32 bytes of data:
Reply from 10.16.108.2: TTL expired in transit.
Reply from 10.16.108.2: TTL expired in transit.
Reply from 10.16.108.2: TTL expired in transit.
Reply from 10.16.108.2: TTL expired in transit.

When i said I had the same setup on the 10.16.198.0/22, there is one diffrence. The switch there is a 2626 Provision Procuve and the other on 10.16.160.0/22 is a 2620 ArubaOS (latest 16 version) Procurve.

As I see it there is nothing I could do the Cisco to fix this. Seems like the Procurve is the one doing this.