[c-nsp] ASR-1K problem with old PPPoE clients.

Ariel Weher aweher at gmail.com
Mon Apr 23 13:23:10 EDT 2012


Hi folks!

I'm facing some problems with our brand new asr 1002's.

We're having some issues with the clients being migrated from the old
7200-NPEG1 to the new BRAS with the following symptoms:

The client (mostly old OSes like W2K and very old linux like debian potato)
connects successfully, got an IP address and all of the PPP parameters,
there is no traffic between the user and the net. Other users connects to
the same BRAS without problems.

This is a brief of our probes while using icmp to test the client
reachability:

BRAS -> Client [OK]
BRAS -> Client (Source interface other than Loopback) [OK]
Client -> BRAS (Any interface) [FAIL]
Client -> any IP [FAIL]
Any other host in the network -> Client [FAIL]

The routes exists in the tables of all of our routers and

Any host of the net -> BRAS (Any local interface) [OK]

The configs are very basic, like

<asr-config>
ip name-server 8.8.8.8
ip name-server 8.8.4.4
!
bba-group pppoe PPPoE
 virtual-template 1
 sessions per-mac limit 1
 sessions per-vlan limit 16000
 sessions per-mac throttle 3 30 120
 sessions auto cleanup
!
interface Loopback0
 description Services
 ip address 200.3.M.N 255.255.255.255
 ip ospf 1 area 0
!
interface Loopback110
 description OSPF Sum 190.X.Y.0/24
 ip address 190.X.Y.1 255.255.255.255
 ip ospf 1 area 400
!
! Multiple (50+) VLAN subif's like this
!
interface GigabitEthernet0/0/2.60
 description PPPoEoVLAN 60
 encapsulation dot1Q 60
 pppoe enable group PPPoE
!
interface Virtual-Template1
 ip unnumbered Loopback0
 peer default ip address pool PPPoE-POOL
 ppp authentication pap
!
router ospf 1
 router-id 200.3.M.N
 log-adjacency-changes detail
 auto-cost reference-bandwidth 10000
 area 400 range 190.X.Y.0 255.255.255.0
!
ip local pool PPPoE-POOL 190.X.Y.2 190.X.Y.254
</asr config>

And the show version:

Cisco IOS Software, IOS-XE Software (PPC_LINUX_IOSD-ADVENTERPRISEK9-M),
Version 15.1(2)S, RELEASE SOFTWARE (fc1)
Technical Support: http://www.cisco.com/techsupport
Copyright (c) 1986-2011 by Cisco Systems, Inc.
Compiled Thu 24-Mar-11 23:29 by mcpre

There are several clients connected to the router and a very little part of
them doesn't get the service.

sh pppoe sum
    PTA  : Locally terminated sessions
    FWDED: Forwarded sessions
    TRANS: All other sessions (in transient state)

                                TOTAL     PTA   FWDED   TRANS
TOTAL                            4221    4220       0       1
GigabitEthernet0/0/2         4221    4220       0       1

The TAC has no responses to give about this problem, and it's happening in
all of our new ASR. If you have any experience about an incident like this,
please drop me a few lines.

AFAIK there is no distinction about the L2 equipment serving the customers.
If I pass the whole VLAN to our old BRAS all of the clients works fine in
the 7200 with the same config.

All of your suggestions are welcome.

Thanks for your time.

Regards.


More information about the cisco-nsp mailing list