Point to Point T1 with Cisco 1720s and 827 DSL router

The 827 Provides internet access to the 10.143.214.X domain. It plugs into 3 x 24 port hubs in order to provide the location with internet access. The local 1720 is connected into one of the hubs, then into the T1 port. 10.143.214.209 is the LAN IP address of the local 1720, not a subnet itself (or at least it shouldn't be).  192.168.1.1 Is the IP I assigned to the Serial0 interface for communication over the T1 line between 1720s.

Would I need to add:

ip route 192.168.1.1 255.255.255.255 10.143.214.209

to the 827?

Here is the config for the remote 1720:

service timestamps debug uptime
service timestamps log uptime
service password-encryption
!
hostname remote
!
enable password xxx
!
no ip name-server
!
ip subnet-zero
no ip domain-lookup
ip routing
!
interface FastEthernet 0
no shutdown
description connected to local_lan
ip address 206.95.134.1 255.255.255.0
keepalive 10
!
interface Serial 0
no shutdown
description connected to T1-1720
service-module t1 clock source line
service-module t1 data-coding normal
service-module t1 remote-loopback full
service-module t1 framing esf
service-module t1 linecode b8zs
service-module t1 lbo none
service-module t1 remote-alarm-enable
ip address 192.168.1.2 255.255.255.252
encapsulation ppp
!
router rip
version 2
network 206.95.134.0
network 192.168.1.0
network 206.95.134.0
no auto-summary
!
!
ip classless

ip route 0.0.0.0 0.0.0.0 206.95.134.110.143.214.100
no ip http server
snmp-server community public RO
no snmp-server location
no snmp-server contact
!
line console 0
 exec-timeout 0 0
 password xxx
 login
!
line vty 0 4
 password xxx
 login
!

end





This remote location is operating on the 206.95.134.X range of IPs. For the life of mee I can't see anything there that would be preventing communication. Any ideas?
 

Now that I see the other 1720, what you are going to need in the 827 is the route to the second subnet on the other side.

You already have the default route to the main subnet, but that subnet has no idea you have another one at the end of the 192.168.1.x WAN.


Add to the 827.

ip route 206.95.134.0 255.255.255.0 192.168.1.2

-Mark

Opps, make that:

ip route  206.95.134.1 255.255.255.0 10.143.214.209.

then on the local 1720, add the above route:

ip route 206.95.134.0 255.255.255.0 192.168.1.1


Sorry about the confusion.  You need to tell it how to get to the remote location.  It knows how to get back with your 0.0.0.0 route.

-Mark

Two questions/observations...

Shouldn't the IP mask be 255.255.255.255 instead of  .0?

and on the local, the route should aim to .2, not .1 since .1 is the local 1720.


Everything looks fine to you on the remote 1720?

Well, after these changes, I telneted into the "local" 1720, and was able to ping a PC on the remote network! I will verify that the internet is working fine tomorrow when I am on-site.

Should this allow access to network shares now, or  do I need additional settings?

No, this add with let all IP traffic flow because you don't have any ACL's between them.

You want to route the whole 206.95.134.x to the remote office, that is why you use the .0.

You can route to the local, or the remote IP.  Never noticed a difference.  You could just as easily route it do the S0 interface.

I see you have RIP enable, so I don't know why we have to use the routes.  But if you use EIGRP, you would have alot more success with the auto routes.

Then on the original 827, you could add 'redistribute static' and all the routes will get delived to all the routers (Subnets) defined via eigrp.

-Mark

If you enable RIP v2 on the 827 (as you have on at least 1 of the 1700's), it should help.

Once you start running multiple routers/sites, dynamic protocols make so much more sense.

RIPv2 works if you wanna stay with that. EIGRP has much faster convergence if you wanna go that route.

The 827 should still probably have a default route and then just redistribute the default route into the IGP or generate default route for the rest of the internal network, like this:

router827:
router rip
 version 2
 network 10.143.214.0
 default-information originate
 passive-interface BVI1
no auto-summary
...
or with the redist static:
router rip
 version 2
 network 10.143.214.0
 passive-interface BVI1
 redistribute static
no auto-summary
...

Basically you need to do the same thing and make sure that this is all kosher on the remote 1720 as well.

Verify that all is good by doing a sh ip route on any of the three routers and you should see all your internal subnets and a default route heard from the 827.

this will also mean that you no longer need any static routes in your network except for on the 827.

Ok, I am currently on-site. Network shares on the remote site are not browseable, but are accessible via share name and/or IP address... typical win98 behavior it seems...

The only issu remaining is that the internet traffic seems to hang for 30-60 seconds before resuming.  Everything will work fine for a few minutes and then hang again, almost like if the connection was dropped.

Is there some kind of keep-alive command I should use???

Checked serial interface on both ends and the local shows a few errors, as show below.

Serial0 is up, line protocol is up
  Hardware is PQUICC with Fractional T1 CSU/DSU
  Description: connected to sopp_ford-chevy
  Internet address is 192.168.1.2/30
  MTU 1500 bytes, BW 1544 Kbit, DLY 20000 usec,
     reliability 255/255, txload 1/255, rxload 26/255
  Encapsulation PPP, loopback not set
  Keepalive set (10 sec)
  LCP Open
  Open: IPCP, CDPCP
  Last input 00:00:01, output 00:00:00, output hang never
  Last clearing of "show interface" counters 3d22h
  Input queue: 0/75/0 (size/max/drops); Total output drops: 0
  Queueing strategy: weighted fair
  Output queue: 0/1000/64/0 (size/max total/threshold/drops)
     Conversations  0/6/256 (active/max active/max total)
     Reserved Conversations 0/0 (allocated/max allocated)
  5 minute input rate 160000 bits/sec, 9 packets/sec
  5 minute output rate 0 bits/sec, 0 packets/sec
     294016 packets input, 275226854 bytes, 0 no buffer
     Received 0 broadcasts, 0 runts, 0 giants, 0 throttles
     72 input errors, 19 CRC, 36 frame, 0 overrun, 0 ignored, 17 abort
     239104 packets output, 88037737 bytes, 0 underruns
Serial0 is up, line protocol is up
  Hardware is PQUICC with Fractional T1 CSU/DSU
  Description: connected to sopp_ford-chevy
  Internet address is 192.168.1.2/30
  MTU 1500 bytes, BW 1544 Kbit, DLY 20000 usec,
     reliability 255/255, txload 1/255, rxload 26/255
  Encapsulation PPP, loopback not set
  Keepalive set (10 sec)
  LCP Open
  Open: IPCP, CDPCP
  Last input 00:00:01, output 00:00:00, output hang never
  Last clearing of "show interface" counters 3d22h
  Input queue: 0/75/0 (size/max/drops); Total output drops: 0
  Queueing strategy: weighted fair
  Output queue: 0/1000/64/0 (size/max total/threshold/drops)
     Conversations  0/6/256 (active/max active/max total)
     Reserved Conversations 0/0 (allocated/max allocated)
  5 minute input rate 160000 bits/sec, 9 packets/sec
  5 minute output rate 0 bits/sec, 0 packets/sec
     294016 packets input, 275226854 bytes, 0 no buffer
     Received 0 broadcasts, 0 runts, 0 giants, 0 throttles
     72 input errors, 19 CRC, 36 frame, 0 overrun, 0 ignored, 17 abort
     239104 packets output, 88037737 bytes, 0 underruns

I did a clear counters after I posted and the errors have not come back. Shares are still not browseable but can be mapped via IP (not a problem tho).

The Internet connection seems to have stabilized for now. if they come back I will post back.

Do you have a WINS server?  They help alot when browsing over a WAN using NT4.0.

If you have a mixed enviroment, use WINS.

If you just have Win2k Servers and don't want to use WINS, make sure that your DNS server will auto-update and all the clients have the Win2k DNS server as their primary.

-Mark