Hi!
I have a trouble with VRF (well, I think that it's with VRF) and I can't understand why this trouble occurs and how to solve it.
Here is a brief description.
I have 3 different models of Cisco Routers on 3 different sites. The models are:
Cisco 2911 IOS UNIVERSALK9-M v. 15.2(4)M2
Cisco 1941 IOS UNIVERSALK9-M v.15.3(3)M3
Cisco 2811 IOS C2800NM-ADVENTERPRISEK9-M v.15.1(3)T3
Usual connection scheme you may find in attached file.
And I have the same trouble on all this routers. Here is short description.
I use router on a stick and I have a few dot1q sub-interfaces on my routers.
Also I use VRF for separating routing tables from different ISPs and from common routing table.
Usually all works fine, but sometimes one of LANs lose connection with other networks and Internet. However, I can ping computers in affected LAN from Router and I can ping router's sub interface of the affected LAN too.
All router tables look fine, while this kind of trouble occur, and also I can see fresh NAT translations from affected LAN, but there is no traffic from and to affected LAN.
Usually I issue «shutdown» and then «no shutdown» on sub interface to temporarily solve this problem. But, I can't understand the general cause.
Additional difficulty is that I can't get much traces or make some tests to understand the problem when it occurs, because users in affected LAN need access to Internet and other LANs. So I just issue «shut – no shut». My tries to repeat problem in lab environment are still unsuccessful.
So now I am asking for help in community. May be somebody encountered with this kind of problem and can give me a good peace of advice.
In attachments you may find simple scheme and partial router config.
I can provide additional information if need.
