LFIB AND FIB

[akachroo123]

How LFIB and FIB are generated,source?

[spremkumar]

hi

Enabling CEF gives you FIB (fowarding information base) which is very much similar to routing table .

It does sits in normlly in ur data place.

LFIB (Label Forwarding Info Base) this is derived from both LIB and FIB.

Again your LIB(Label Information Base) is derived when you enable tag switching on the interfaces.

This again resides in Control plane.

but your LFIB sits out in Data plane where you will see data being forwarded based on upper label stack and which is basically used to insert or swap labels based on the destination prefix.

hope this may help u out..

regds

[akachroo123]

So you mean to say that these 3 tables are their in every PE and CORE routers(MPLS NETWORK).

And how these tables are related/interlinked with normal routing tables.Is MPLS only compatiable with BGP.

[Harold Ritter]

These 3 tables would indeed be present on all P (core routers) and PE routers.

One thing to add to spremkumar excellent explanation. The LIB is populated by LDP (or TDP). You don't necesseraly need LDP to setup LSPs. MPLS TE can also setup LSPs, which means the LFIB is not only populated from the FIB and LIB but from MPLS TE as well.

As far as your question on how these tables are related to the normal routing table (AKA the RIB), the FIB is created from the RIB and is basically a subset of it.

MPLS is compatible with all routing protocols. The only exception is that LDP will allocate labels for all prefixes in the RIB except for BGP learnt prefixes. The reason for that is that the BGP next hop is usually learn via IGP or is directly connected, which in both cases defeats the purpose of allocating a label to these prefixes since a label is already allocated for the recursive nexthop.

Hope this helps,

[akachroo123]

One more thing,when traffic starts flowing from CE through PE from the first time,how all these tables start building up,can you explian me the steps.

One more thing does MPLS supports load balancing.

[attrgautam]

When traffic from the CE comes to the PE router , there will be a label for the remote end prefix. So when the router forwards the packet ,it will tag the label for the remote end packet and switch it to the corresponding interface.

MPLS does support load balancing by giving the same label for a prefix obtained across all links

[mounir.mohamed]

There is a tow situations could effect on the packet when received from CE on PE

The fist one is if it’s normal ip packet the router perform an normal routing lookup to find the destination address if it’s found and the outgoing interface is MPLS enabled the router attach the packet with a new label the core routers swap between labels

Second situation if the packer come from CE to PE and the customer interface attached with MPLS-VPN (VRF) then the router attached the packer with 2 labels one for the egress router and the second one for the prefix, BGP used in this situation to advertise both labels.