Question Locked

problems with a stub network on Frame Relay network

0 helpful
19 replies
Closed to new repliesThis discussion is closed. It stays readable, but it takes no new replies.

Working with a client that has a private frame-relay network. While there are no issues with the frame-relay network, they are trying to connect the frame-relay network to the Internet via a stub network. The stub is off a 3640 at one of the sites using an ethernet connection from the 3640 to the perimeter router, then off to the internet. Everyone on the frame-relay network [all subnets] can ping the ethernet interface on the 3640, but they can not ping the DSL perimeter router. eg packets are dropped at the 3640 ethernet interface that connects to their DSL router [the stub network].

Here's a ip route table [RIP] from the 3640: Router#sh ip route
Codes: C - connected, S - static, R - RIP, M - mobile, B - BGP
D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
E1 - OSPF external type 1, E2 - OSPF external type 2
i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
ia - IS-IS inter area, * - candidate default, U - per-user static route
o - ODR, P - periodic downloaded static route

Gateway of last resort is 192.168.X.X to network 0.0.0.0

172.16.0.0/24 is subnetted, 3 subnets
R 172.16.40.0 [120/1] via 172.16.20.2, 00:00:08, Serial0/0
R 172.16.30.0 [120/1] via 172.16.20.1, 00:00:09, Serial0/0
C 172.16.20.0 is directly connected, Serial0/0
173.55.0.0/16 is variably subnetted, 2 subnets, 2 masks
R 173.55.0.0/16 [120/3] via 172.16.20.2, 00:00:08, Serial0/0
R 173.55.107.0/24 [120/1] via 192.168.X.X, 00:00:16, Ethernet0/0
C 192.168.X.X/24 is directly connected, Ethernet0/0
S* 0.0.0.0/0 [1/0] via 192.168.X.X
is directly connected, Ethernet0/0

The 192.168.X.X directly connected to ethernet 0/0 is the stub network off to the DSL router. The DSL router has the 192.168.X.X subnet.

Note: you can ping yahoo from the 3640 ...

Here's the ip route table [RIP] from one of the sites :
Gateway of last resort is not set

172.16.0.0/24 is subnetted, 3 subnets
R 172.16.40.0 [120/1] via 172.16.20.2, 00:00:03, Serial0/0
C 172.16.30.0 is directly connected, FastEthernet0/0
C 172.16.20.0 is directly connected, Serial0/0
R 173.55.0.0/16 [120/3] via 172.16.20.2, 00:00:03, Serial0/0
R 192.168.X.X/24 [120/1] via 172.16.20.2, 00:00:03, Serial0/0
Router#

While the gateway hasn't been set at the site they can ping the 3640 ethernet interface via the frame-relay network.

They advised me they set the gateway to the DSL perimeter router 192.168.x.x [1 hop away] eg o.o.o.o = 192.168.x.x and that didn't work.

IP classless is running on all routers.

Any ideas why the stub network will not allow icmp ping from frame-relay sites to the perimeter router ?

note no ACLs on any router and no firewalls block icmp traffic. any host on any of the frame-relay sites can ping the ethernet interface on the 3640 which is the stub network and where the DSL perimeter router connect to. It appears that when icmp ping gets to the 3640 ethernet interface, it doesn't forward them to the DSL router rather drops the packets.

19 replies

by netman1958

Does the DSL router have routes to the 172.16.X.X networks? If it just has a defualt route then it must point to the ISP's network which will drop traffic destined for 172.16.X.X.

From the 3640 itself a ping will be sourced from 192.168.X.X which is connected to the DSL router so it will have a route back to the 3640.

If you ping from one of the 172.16.X.X devices, it's source IP will be 172.16.X.X and when the ping reaches the DSL router, if all it has is one default route and no specific routes back to the 172.16.X.X net, it will try to send them through the ISP.

by cg it

static routes created to the 172.16.X.X frame-relay network subnets....

but I'll have then run a ip route command to verify that in the DSL router that the 172.16.X.X networks show up in the route table.

Thanks Netman!!

by robo_dev

I'm no Cisco guru, and it's hard to understand the topology exactly without a Visio diagram, but are you using RIP for the stub router?

Most of the config examples I see are for EIGRP or OSPF for stub routers....or am I confused?

by cg it

I think whoever first set it up probably didn't know Eigrp or OSPF.

Here's a diagram for you

Frame-relay switch with 8 Async ports at HQ

each site connects to 1 of the 8 Async ports and used frame-relay encapsulation. There is a frame-relay circuit between each site. The frame-relay switch simply moves frames around between each site. it's a completely closed network with no internet access.

someone go the idea at one of the sites to setup internet access via a DSL line they had, so they setup a stub network off their site router [which is a 3640] by running a cat 5 from an unused ethernet port on the 3640 to the DSL router/modem.

by netman1958

Is 172.16.20.2 assigned to the 3640? The reason I ask is because the 3640 has routes to a couple of subnets with a next hop of 172.16.20.2:
R 172.16.40.0 [120/1] via 172.16.20.2, 00:00:08, Serial0/0
R 173.55.0.0/16 [120/3] via 172.16.20.2, 00:00:08, Serial0/0

While the other router who's routing table you posted has a route to 192.168.X.X with a next hop of 172.16.20.2:
R 192.168.X.X/24 [120/1] via 172.16.20.2, 00:00:03, Serial0/0

From that I gather that the 3640's serial0 address isn't 172.16.20.2 and that the other router isn't sending traffic for 192.168.X.X directly to the 3640.

by cg it

173.55.107.1 255.255.255.255 * 10 eth1
173.55.107.1 255.255.255.255 * 11 ipsec0
172.16.20.0 255.255.255.0 192.168.X.X 10 eth0
192.168.X.0 255.255.255.0 * 0 eth0
172.16.30.0 255.255.255.0 192.168.X.X 10 eth0
173.55.107.0 255.255.255.0 * 0 eth1
173.55.107.0 255.255.255.0 * 0 ipsec0
172.16.40.0 255.255.255.0 192.168.X.X 10 eth0
default 0.0.0.0 173.55.107.1 40 eth0

So it looks like they have all the frame-relay networks using the 192.168.X.X as the default gateway...

well it does give the path out... I asked em to try and ping their ISP name servers and they get destination unreachable from the 172.16.20.2 interface....

by cg it

well its on another router at another site I'm told...

While they didn't give me the route table for that router I'll bet it's not right...

by netman1958

Can you get a static route configured on one of the hub routers like :
ip route 0.0.0.0 0.0.0.0 X.X.X.X
where X.X.X.X is the serial0 IP of the 3640.

then try to ping the dsl router from that hub router. If it doesn't ping, try a traceroute from the hub to a IP on the Internet and see where it starts timing out.

by cg it

This is the 172.16.20.2 network and it's here thats the problem...

Gateway of last resort is not set

172.16.0.0/24 is subnetted, 3 subnets
C 172.16.40.0 is directly connected, FastEthernet0/0
R 172.16.30.0 [120/1] via 172.16.20.1, 00:00:15, Serial0/1
C 172.16.20.0 is directly connected, Serial0/1
R 173.55.0.0/16 [120/2] via 172.16.20.3, 00:00:07, Serial0/1
R 192.168.X.X/24 [120/1] via 172.16.20.3, 00:00:07, Serial0/1
Router#

the S0/1 interface address is the 172.16.20.2 address.

by cg it

shows
1=172.16.30.1
2=172.16.20.2
3=172.16.20.3
4=192.168.X.X

but I think the other network, the 172.16.20.2 is not configured right with the correct route tables.

and I think their frame-relay switch is simply a frame-relay switch with simple circuits via DLCIs between all their sites.

So that complicates matters a bit.

by netman1958

If this router is 172.16.20.2, it has a route to 192.168.X.X via 172.16.20.3 that it says is 1 hop away [120/1]:
R 192.168.X.X/24 [120/1] via 172.16.20.3, 00:00:07, Serial0/1

Yet the other router in your first post has a route to 192.168.X.X via 172.16.20.2 that it thinks is 1 hop away [120/1]:
R 192.168.X.X/24 [120/1] via 172.16.20.2, 00:00:03, Serial0/0

Makes me think that 172.16.20.3 is the 3640 and that all of the other routers should have a route like this:
R 192.168.X.X/24 [120/1] via 172.16.20.3, 00:00:03, Serial0/0

Unless split-horizon has been disabled on some of the routers, they shouldn't advertise a route to 192.168.X.X back out the same interface (serial0) that they received it on so it's confusing. Scratching my head here.

by cg it

what I think is that each of the site routers serial interfaces has an address for their respective subnets and that to connect to other sites, there is a virtual circuit via DLCI designations. and the frame-relay switch interfaces dont' have addresses. It's a simple frame switch. that's the only explanation I could come up with where the serial interfaces seem to connect with each other. Something like this

DLCI 111 to DLCI 222
DLCI 222 to DLCI 112
DLCI 113 to DLCI 333
DLCI 333 to DLCI 133
etc...

where DLCI 111=172.16.20.1 and DLCI 222 =172.16.20.2

then DLCI333=172.16.20.3

so there's a circuit to each network...pretty simple setup frame-relay network but since they say they've never had problems it works ok...

by cg it

really appreciate the help... not often there's Cisco stuff here...

by netman1958

I actually had an issue similar to this once when I added some additional subnets behind an internet router. You posted traceroute results from 172.16.40.X that looked like this:
1=172.16.30.1
2=172.16.20.2
3=172.16.20.3
4=192.168.X.X

What IP were they trying to ping when they got that and is hop 4 192.168.X.X the IP of the DSL router?

by cg it

I believe the trace route was run on a server on the 172.16.30.1 subnet not the 172.16.40.1

I think hop 3 is the 3640 router

not sure what hop 2 router is

but the 172.16.30.1 router is the site router where the server is.

Just thinking here, the path probably bounces around the frame circuits since their virtual circuits until it gets to the 3640 router which then knows to forward to the 0.0.0.0 gateway.

I can see network traffic really chewing up bandwidth if they ever get to the internet because packets are bouncing around frame-relay virtual circuits.

A ping request from a server on the 172.16.30.1 subnet which is a site doesn't go directly to the 3640 router and out, it first goes to another site which is probably the 172.16.20.2 router. that router then sends it to the 3640 router and then out to the DSL router.

lol like ping pong because of the virtual circuits. once it goes through the circuit, it then gets sent to the 3640...

by netman1958

Sorry, I had to cut my last post short because I had to leave the office for something. I am home now and I think I have it figured out.

Like I started to explain in the previous post that I had to cut short, I once had a similar issue when I added some additional subnets behind an internet gateway. I configured the extra routes on our frame-relay hub router and the new spoke subnets could communicate with the hub subnet and even ping the internet gateway router. But they couldn't get to anything on the internet.

After lots of trouble-shooting and pulling my hair out it finally hit me that I hadn't configured the internet gateway router to perform NAT on the new subnets. Once I added the new subnets to the NAT access-list everything worked beautifully.

Since the traceroute from the 172.16.30.X subnet does reach the DSL router and because you can ping yahoo from the 3640 router, I'll bet that the DSL router is configured to NAT traffic from 192.168.X.X only. Take a look at it's config and if that is the case, reconfigure it to include the 172.16.X.X subnets in NAT.

by cg it

Thanks for all the help Netman..I'd be pulling my hair out for probably ever...

This discussion is closed to new replies.