Monitoring NetFlow from a PPP interface...

I'm  having a little trouble following your set up. Can you give me an idea of the topology and what you want to see?

Yeah, sorry about that.  Here is a quick image of the setup (the important part, at least).  All I'm trying to monitor is the T1 at the remote site.

http://grab.by/3Lqv

So, originally when adding nodes to NPM, I would select the router, and told it to monitor the T1 1/1 interface.  Doing this allowed me to see stats like error counters, etc, thanks to SNMP.  

The problem with that, though, is that the T1 1/1 interface doesn't have an IP address - the logical PPP1 interface has the IP.  

When I configure NetFlow on the router, I have to tell it where to source the NetFlow traffic from.  I can only source the traffic from an interface with an IP address (ie: must be a layer3 interface).  So I tell it to source from the PPP1 interface.

Now I am monitoring 2 interfaces, although they are really one in the same.  Within NPM, I need the T1 1/1 interface because it holds SNMP statistics.  Within NetFlow, I need PPP1 since it sends NetFlow traffic.

Does that make more sense?

Edit: As a side note - the reason this is annoying is because I am now effectively doubling the total amount of interfaces being monitored.  Also, it's a little more cluttered since I don't have one interface with ALL information.

can you give me a config snippet I can test? I'll make up the Ip Addresses

Config snippet from what? The router?

Yep - for the PPP and T-1 IF's

interface t1 1/1

  description MPLS-T1

  fdl none

  tdm-group 1 timeslots 1-24 speed 64

  no shutdown

interface ppp 1

  ip address  X.X.X.X  255.255.255.252

  ip flow ingress

  ip flow egress

  media-gateway ip primary

  max-reserved-bandwidth 100

  qos-policy out QOS_MAP

  no shutdown

  cross-connect 1 t1 1/1 1 ppp 1

Is that an Adtran box? It's so close to Cisco-speak it took me a second to realize it's not...

Can't you just put a loopback address on the router and add the device to NPM to be monitored from the looback ip.  You then want to export netflow data from the loopback address also.  You need to have config with some lines like these:

ip flow-export source Loopback0
ip flow-export version 9
ip flow-export destination <put your NPM IP here> 2055

logging trap debugging
logging facility syslog
logging source-interface Loopback0

logging <put your NPM IP here>

snmp-server ifindex persist
snmp-server location
snmp-server contact
snmp-server enable traps snmp authentication linkdown linkup coldstart warmstart
snmp-server enable traps tty
snmp-server enable traps bgp
snmp-server enable traps config
snmp-server enable traps syslog
snmp ifmib ifindex persist

this way you monitor thought the loopback and get syslog, traps, and netflow from the same IP.

Haha good catch! 

It is, and yes it's nearly identical :)

Thanks for the replies - I will experiment with loopback and see if that does the trick!

If I understand the question correctly, I don't think the loopback configuration is going to help. It sounds like the problem is that Tyler doesn't want to view the physical interface and the logical interface on separate pages.

This doesn't seem to me like a NTA issue per se; it's more the consequence of the fact that the physical and logical interfaces have different SNMP ifIndexes, so Orion puts them on different pages.

I'll be interested to see if Solarwinds can come up with a way to combine multiple interfaces onto a single page; this would be useful not only in a case like this but in cases like Multilink PPP interfaces, dialer interfaces, or SVIs (aka VLAN interfaces) and their associated physical members.