[c-nsp] Cisco 1720 and ip cef

Oliver Boehmer (oboehmer) oboehmer at cisco.com
Sat Sep 25 05:05:55 EDT 2004


Hi Nigel,

you are using only one Layer3 interface (Dialer1), so CEF only sees one
link, resulting in CEF load-balancing being completely taken out of the
picture here. 
As far as I know there are two ways to achieve load-sharing here:

1) use multilink PPP on the dialer interface to create on Layer3
interface and have PPP bundle the two members at Layer2 (you might want
to disable multilink fragmentation to increase performance). This
obviously needs to be supported on the concentrator across the DSL
cloud:

int dialer1
 [...]
 ppp multilink
 no ppp multilink fragmentation

2) use two Dialer interfaces (i.e. create two dialer pools) and thus use
CEF load-balancing via two Layer3 interfaces.

interface ATM0.3 point-to-point
 pvc 8/35
  dialer pool-member 1
 !
interface ATM1.3 point-to-point
 pvc 8/35
  dialer pool-member 2
 !

int Dialer1
 ...
 dialer pool 1
int Dialer2
 ...
 dialer pool 2
!
ip route 0.0.0.0 0.0.0.0 Dialer1
ip route 0.0.0.0 0.0.0.0 Dialer2

Hope this helps,

	oli

cisco-nsp-bounces at puck.nether.net <> wrote on Saturday, September 25,
2004 3:46 AM:

> Hi Rodney,
> 
> As requested:
> 
> a) Added ip cef accounting per-prefix (as suggested in (d) -
> the actual
> command you mentioned was not available)
> 
> version 12.3
> <snip>
> !
> !
> ip cef
> ip cef accounting per-prefix
> !
> !
> !
> !
> interface ATM0
>  no ip address
>  no atm ilmi-keepalive
>  dsl operating-mode itu-dmt
> !
> interface ATM0.3 point-to-point
>  pvc 8/35
>   ubr 256
>   encapsulation aal5mux ppp dialer
>   dialer pool-member 3
>  !
> !
> interface ATM1
>  no ip address
>  no atm ilmi-keepalive
>  dsl operating-mode itu-dmt
> !
> interface ATM1.3 point-to-point
>  pvc 8/35
>   ubr 256
>   encapsulation aal5mux ppp dialer
>   dialer pool-member 3
>  !
> !
> interface FastEthernet0
>  ip address xxx
>  speed auto
>  no keepalive
> !
> interface Dialer3
>  bandwidth 512000
>  ip address negotiated
>  ip load-sharing per-packet
>  encapsulation ppp
>  dialer pool 3
>  ppp chap hostname xxx
>  ppp chap password xxx
> !
> ip classless
> ip route 0.0.0.0 0.0.0.0 Dialer3
> <snip>
> end
> 
> b)
> Gateway of last resort is 0.0.0.0 to network 0.0.0.0
> 
>      172.16.0.0/16 is variably subnetted, 3 subnets, 2 masks
> C       172.16.254.3/32 is directly connected, Dialer3
> C       172.16.254.252/32 is directly connected, Dialer3
> C       172.16.3.0/24 is directly connected, FastEthernet0
> S*   0.0.0.0/0 is directly connected, Dialer3
> 
> 
> c)
> Router#sho ip cef
> Prefix              Next Hop             Interface
> 0.0.0.0/0           attached             Dialer3
> 0.0.0.0/32          receive
> 172.16.3.0/24       attached             FastEthernet0 172.16.3.0/32 
> receive 172.16.3.1/32       172.16.3.1           FastEthernet0
> 172.16.3.11/32      172.16.3.11          FastEthernet0
> 172.16.3.12/32      172.16.3.12          FastEthernet0
> 172.16.3.13/32      172.16.3.13          FastEthernet0
> 172.16.3.14/32      172.16.3.14          FastEthernet0
> 172.16.3.15/32      172.16.3.15          FastEthernet0
> 172.16.3.20/32      172.16.3.20          FastEthernet0
> 172.16.3.31/32      172.16.3.31          FastEthernet0
> 172.16.3.32/32      172.16.3.32          FastEthernet0
> 172.16.3.53/32      172.16.3.53          FastEthernet0
> 172.16.3.54/32      172.16.3.54          FastEthernet0
> 172.16.3.55/32      172.16.3.55          FastEthernet0
> 172.16.3.71/32      172.16.3.71          FastEthernet0
> 172.16.3.82/32      172.16.3.82          FastEthernet0
> 172.16.3.83/32      172.16.3.83          FastEthernet0
> 172.16.3.84/32      172.16.3.84          FastEthernet0
> 172.16.3.85/32      172.16.3.85          FastEthernet0
> 172.16.3.92/32      172.16.3.92          FastEthernet0
> 172.16.3.102/32     172.16.3.102         FastEthernet0
> 172.16.3.103/32     172.16.3.103         FastEthernet0
> 172.16.3.108/32     172.16.3.108         FastEthernet0
> 172.16.3.113/32     172.16.3.113         FastEthernet0
> 172.16.3.114/32     172.16.3.114         FastEthernet0
> 172.16.3.115/32     172.16.3.115         FastEthernet0
> 172.16.3.116/32     172.16.3.116         FastEthernet0
> 172.16.3.118/32     172.16.3.118         FastEthernet0
> 172.16.3.120/32     172.16.3.120         FastEthernet0
> 172.16.3.131/32     receive 172.16.3.161/32     172.16.3.161        
> FastEthernet0 172.16.3.162/32     172.16.3.162         FastEthernet0
> 172.16.3.163/32     172.16.3.163         FastEthernet0
> 172.16.3.164/32     172.16.3.164         FastEthernet0
> 172.16.3.165/32     172.16.3.165         FastEthernet0
> 172.16.3.166/32     172.16.3.166         FastEthernet0
> 172.16.3.167/32     172.16.3.167         FastEthernet0
> 172.16.3.168/32     172.16.3.168         FastEthernet0
> 172.16.3.169/32     172.16.3.169         FastEthernet0
> 172.16.3.170/32     172.16.3.170         FastEthernet0
> 172.16.3.171/32     172.16.3.171         FastEthernet0
> 172.16.3.172/32     172.16.3.172         FastEthernet0
> 172.16.3.173/32     172.16.3.173         FastEthernet0
> 172.16.3.174/32     172.16.3.174         FastEthernet0
> 172.16.3.175/32     172.16.3.175         FastEthernet0
> 172.16.3.181/32     172.16.3.181         FastEthernet0
> 172.16.3.182/32     172.16.3.182         FastEthernet0
> 172.16.3.183/32     172.16.3.183         FastEthernet0
> 172.16.3.184/32     172.16.3.184         FastEthernet0
> 172.16.3.255/32     receive 172.16.254.3/32     receive
> 172.16.254.252/32   attached             Virtual-Access3
>                     attached             Virtual-Access2
> Prefix              Next Hop             Interface
> 224.0.0.0/4         drop
> 224.0.0.0/24        receive
> 255.255.255.255/32  receive
> 
> d)
> 
> Router(config)#ip cef accounting ?
>   non-recursive  Enable accounting for traffic through non-recursive
>   prefixes per-prefix     Enable per prefix accounting
>   prefix-length  Enable prefix length accounting
> 
> Router(config)#^Z
> Router#sh ip cef 0.0.0.0 ?
>   A.B.C.D  prefix mask
>   detail   Display full information
>   |        Output modifiers
>   <cr>
> 
> Router#sh ip cef 0.0.0.0
> 0.0.0.0/32, version 0, epoch 0, receive
> Router#sh ip cef 0.0.0.0 0.0.0.0
> 0.0.0.0/0, version 7, epoch 0, attached
> 0 packets, 0 bytes
>   via Dialer3, 0 dependencies
>     valid punt adjacency
> 
> Thanks again.
> 
> Nigel
> 
> 
> 
> -----Original Message-----
> From: Rodney Dunn [mailto:rodunn at cisco.com]
> Sent: Friday, September 24, 2004 10:31 PM
> To: Nigel
> Cc: cisco-nsp at puck.nether.net
> Subject: Re: [c-nsp] Cisco 1720 and ip cef
> 
> We need to see:
> 
> a) the configuration
> b) show ip route
> c) show ip cef
> d) if you are using a default route out we
>    need you to enable:
>    ip cef accounting load
>    and do 'sh ip cef 0.0.0.0 int' a couple of times.
> 
> Rodney
> 
> 
> On Fri, Sep 24, 2004 at 10:49:53AM +1000, Nigel wrote:
>> Hi All,
>> 
>> I am running into some trouble regarding the cisco 1720 and ip cef
>> load balancing. 
>> 
>> The router has 2 adsl wics in it. Each dsl connection is 1.5M/256.
>> 
>> The problem is that on the outbound path the traffic is only ever
>> going out 1 specific wic. I have tried forcing cef to per packet in
>> order to push traffic out both interfaces, but this still results in
>> traffic only going out the 1 interface still. If I shutdown the
>> interface that is passing outbound traffic the router drops of the
>> network completely. 
>> 
>> The inbound traffic comes down both wics. So far the ios has been
>> upgraded to 12.3(10) and the wic has been swapped out.
> Neither has made a
> difference.
>> 
>> It leaves the actual wic slot on the 1720 or the possibility that the
>> 1720 just can't deal with load balancing.
>> 
>> Here are the outputs from the interfaces:
>> 
>> Router#sho int atm0 | inc rate
>>   Queueing strategy: Per VC Queueing
>>   5 minute input rate 112000 bits/sec, 38 packets/sec
>>   5 minute output rate 0 bits/sec, 0 packets/sec
>> 
>> 
>> Router#sho int atm1 | inc rate
>>   Queueing strategy: Per VC Queueing
>>   5 minute input rate 110000 bits/sec, 37 packets/sec
>>   5 minute output rate 32000 bits/sec, 98 packets/sec
>> 
>> 
>> Any ideas at all?
>> 
>> Thanks in advanced.
>> 
>> Regards,
>> 
>> Nigel
>> 
>> _______________________________________________
>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/
> 
> 
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/



More information about the cisco-nsp mailing list