Netflow packets not getting to inside network

Unanswered Question
May 15th, 2007
User Badges:

I have an external router set up for Netflow with a destination ip address of an internal host which has the netflow Analyzer software on it. I have mirrored the configuration on the inside router which is talking just fine with the Netflow server. I am doing PAT on the outside interface of the firewall with ip address of x.x.x.131, and the outside interface has an ip address of x.x.x.130, and the interface on the router connected to the the outside interface of the firewall has an IP address of x.x.x.129. I have logging set to level 7 and am not seeing the udp port 9996 (netflow information) coming into the firewall. But when i do a

SECRTREXT01# show ip flow export

Flow export v5 is enabled for main cache

Exporting flows to internalsubnet.10.250 (9996)

Exporting using source interface FastEthernet0/1

Version 5 flow records

9780 flows exported in 360 udp datagrams

0 flows failed due to lack of export packet

359 export packets were sent up to process level

0 export packets were dropped due to no fib

0 export packets were dropped due to adjacency issues

0 export packets were dropped due to fragmentation failures

0 export packets were dropped due to encapsulation fixup failures



The interface connecting to the PIX as you see above is F0/1. I had put an access-list on this interface outbound allowing anything and logging but nothing showed for port 9996? insidehost10.250 is being patted to a x.x.x.131. Could this be a routing problem? I have a default route (gateway of last resort) pointing to the upstream LEC router, which is matched only after the destination does not match anything in the x.x.x.128-160 range (/27).

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
Jon Marshall Tue, 05/15/2007 - 23:15
User Badges:
  • Super Blue, 32500 points or more
  • Hall of Fame,

    Founding Member

  • Cisco Designated VIP,

    2017 LAN, WAN

Hi


Your netflow output is saying it exporting flows to "internalsubnet.10.250". But you need to tell your netflow to export flows to your x.x.x.131 address as this is the address the netflow router knows how to get to.


One other point. When you say you have patted the internal server do you mean you have entry on your firewall as such


static (inside,outside) x.x.x.131 internalsubnet.10.250 netmask 255.255.255.255


You will need this.


HTH


Jon

travis.wright Mon, 06/04/2007 - 06:49
User Badges:

Ok thanks a lot... So if i point the netflow to .131 then it will now go in the right direction and hit my firewall, but from there how will it get to the Netflow server without me having to translate 10.250? I am thinking i should set up a static translation to a different global routable IP x.x.x.136 for the Netflow server and point the flows to that ip address. Right now everything going out to the internet is translated as a .131+port so using a different .128/27 ip i am sure will do the trick.. What do you think?

Jon Marshall Mon, 06/04/2007 - 08:52
User Badges:
  • Super Blue, 32500 points or more
  • Hall of Fame,

    Founding Member

  • Cisco Designated VIP,

    2017 LAN, WAN

Hi


I think what you suggest is the best way to achieve what you are trying to do. Use a separate public IP address for the Netflow server and then set up a static translation + the acl rules to allow the netflow traffic through from the router.


Let me know how you get get on


Jon

travis.wright Mon, 06/04/2007 - 09:28
User Badges:

Thanks a lot. I know this will work definitely. I just have to wait for the approval. I know this is changing directions but do you have any suggestions on how i would engineer an architecture where i will have 2 failover pais? They will be PIX 515E's. The top firewalls will be hosting our DMZs and the outside internet of course,. and the bottom pair will be servicing our trusted and restricted segments. The segment design is good but aside from the failover cables i am thinking that the outside interfaces of the bottom firewalls will link directly to the inside interfaces of the top firewalls. However i am thinking that if the bottom or top fails over,.. how will it have to be designed so that the failover pix will communicate with the other primary firewall? I have it designed such that i have an interface on the bottom primary to the top secondary and vice versa. each interface is in a different network of course,.. but how will i do the routing? Since i can only point a default route to 1 IP address. I plan on using stateful failover cables for both UR and FO firewall pairs,.. but any suggestions on how to set up the routing so that the top or bottom of the "sandwich" can still talk to the other half once a failover on the top or bottom occurs? Thanks in advance.

dphills18 Fri, 02/15/2008 - 14:32
User Badges:

did you ever get this to work. we are having the exact same issues.

Actions

This Discussion