08-25-2011 12:02 PM - edited 03-04-2019 01:24 PM
Example of the root problem
Send a ping (echo request) to outside destination sources from Dialer 0. Echo request goes out Dialer 1 interface sourced with Dialer 0s IP address and
gets dropped.
If the echo rquest goes out the Dialer 0 interface, everything works correctly. if only I could get teh infrormation whos source is dialer 0's address to always go out the dialer 0 physical and logical interface.
08-25-2011 01:23 PM
Chris,
your question is very cryptic and if we do not understand what you are asking we cannot try to reply.
Anyway I like to gamble and I am going to attempt a reply.
You connect to an ISP and get 2 different IP addresses with PPPoE to 2 different dialer interfaces.
You route out trying to load balance between the two but not always the traffic comes back.
How do you do the routing to the ISP? How does the ISP knows the routes back to you network? If it does not that is where the problem lies.
Some additional details might be useful.
Fabio
08-25-2011 02:18 PM
Taffic has no problem getting to my router via public IP space. Lets say we have Dialer0 = 1.1.1.1 and Dialer 1 = 2.2.2.2. When I send a ping out since both dialer's have the same cost the icmp echos will alternate interfaces. So what you see in wireshark while wiresharking both interfaces a packet sourced from 1.1.1.1 will alternate out both dialer interfaces. When it goes out Dialer 1 it gets dropped somewhere, when it goes out dialer 0 it get to it's destination fine.
Chris Hercules
Network Engineer / IASO
christopher.hercules@qinetiq-na.com
08-25-2011 03:30 PM
Chris,
The issue is simptomatic of a routing or NAT problem.
Are you using public IP space or private? Is you ISP cooperating with this?
These are the issues you should try to clear.
If you use nat and go out from one interface, coming back from the other won't work.
Is the same ISP on the other end of both PPPoE connections?
I do not believe I have enough information to try to solve the issue here, .... maybe somebody else.
08-26-2011 01:34 PM
Chris,
Are you using CEF 's default load-sharing method (by default it uses per destination method)? You may want to try using per-packet (this will increase CPU and reorders should be expected).
http://www.cisco.com/en/US/products/hw/modules/ps2033/prod_technical_reference09186a00800afeb7.html
Regards.
08-27-2011 07:02 AM
i think this is due to nating issue
please make sure you have a route-map in each nat that match the exit interface
have a look at the bellow link and look at the NATing part once you have it like this you will be able to ping and source the traffic from the exit interface
https://supportforums.cisco.com/docs/DOC-8313
HTH
if helpful Rate
08-29-2011 01:20 PM
thnks for looking into this. I'm not using nat and to make it clear, my next hop is the same address for both interfaces. If you notice the PBRs and routers refer to a different next hop address.
Since both of my interfaces have the same next hop address and
Chris Hercules
Network Engineer / IASO
Supporting Airborne Battle Command PSO
QinetiQ North America
254-532-8321 X 2884
cell: 254-913-8149
chris.hercules@us.army.mil
christopher.hercules@qinetiq-na.com
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: