JunOS mpls forwarding vs bgp-igp.

BGP-IGP VS MPLS-FORWARDING

inet.0 table

This is the primary unicast routing table used by IPv4. It’s where IGPs will resolve next hops.

inet.3 table

This is the routing table populated by MPLS protocols such as RSVP or LDP. A lookup here will result in a label being pushed. inet.3 is used by BGP to resolve BGP next-hops. BGP examines both inet.3 and inet.0 choosing a next-hop based on the lowest preference value. In case of a tie, inet.3 is used.

MPLS traffic-engineering options

You can enable only one of the traffic-engineering statement options at a time

• bgp (default)
• bgp-igp
• bgp-igp-both-ribs
• mpls-forwarding

The default BGP option (traffic-engineering bgp) allows only BGP to use LSPs in its route calculations

bgp-igp

The bgp-igp option causes all inet.3 routes to be moved to the inet.0 routing table. If you configure the traffic-engineering bgp-igp statement or the traffic-engineering bgp-igp-both-ribs statement, high-priority RSVP and LDP routes can supersede IGP routes in the inet.0 routing table. The effect can be noted when we are matching against a protocol in a routing policy. IGP routes might no longer be redistributed since they are no longer the active routes.
LSPs will still be used to resolve BGP next-hops because, BGP can use both tables, it just prefers inet.3 With bgp-igp all your other traffic gets to use the LSP as well not just BGP traffic

Only BGP traffic gets to use the LSP by default so as to support MPLS VPNS, moving everything out of inet.3 breaks your ability to run MPLS VPNs.

root@R6-Junos1> show route 102.102.102.102

inet.0: 21 destinations, 35 routes (21 active, 0 holddown, 0 hidden)
@ = Routing Use Only, # = Forwarding Use Only
+ = Active Route, - = Last Active, * = Both

102.102.102.102/32 *[LDP/9] 00:00:33, metric 1
                    > to 192.168.46.4 via em0.0, Push 28
                    [OSPF/150] 00:00:33, metric 0, tag 0
                    > to 192.168.46.4 via em0.0

bgp-igp-both-ribs

traffic-engineering bgp-igp-both-ribs statement does the same as bgp-igp but instead of moving the contents of inet.3 to inet.0 it copies it over making the labeled path exist in both inet.0 and inet.3 tables. As we have seen with the bgp-igp option, this means high-priority RSVP and LDP routes can supersede IGP routes in the inet.0 routing table. The effect can be noted when we are matching against a protocol in a routing policy. IGP routes might no longer be redistributed since they are no longer the active routes.

root@Router1> show route 2.2.2.2

inet.0: 15 destinations, 20 routes (15 active, 0 holddown, 0 hidden)
+ = Active Route, - = Last Active, * = Both

2.2.2.2/32         *[LDP/9] 00:00:18, metric 1
                    > to 10.10.12.2 via em0.0
                    [IS-IS/18] 00:03:01, metric 10
                    > to 10.10.12.2 via em0.0

inet.3: 4 destinations, 5 routes (4 active, 0 holddown, 0 hidden)
+ = Active Route, - = Last Active, * = Both

2.2.2.2/32         *[LDP/9] 00:00:18, metric 1
                    > to 10.10.12.2 via em0.0

mpls-forwarding

With mpls-forwarding RSVP and LDP routes are used for forwarding but are excluded from route selection. These routes are added to both the inet.0 and inet.3 routing tables. RSVP and LDP routes in the inet.0 routing table are given a low preference when the active route is selected. However, RSVP and LDP routes in the inet.3 routing table are given a normal preference and are therefore used for selecting forwarding next hops.

root@R6-Junos1> show route 102.102.102.102

inet.0: 21 destinations, 35 routes (21 active, 0 holddown, 0 hidden)
@ = Routing Use Only, # = Forwarding Use Only
+ = Active Route, - = Last Active, * = Both

102.102.102.102/32 @[OSPF/150] 00:03:56, metric 0, tag 0
                    > to 192.168.46.4 via em0.0
                    #[LDP/9] 00:00:02, metric 1
                    > to 192.168.46.4 via em0.0, Push 28

inet.3: 14 destinations, 14 routes (14 active, 0 holddown, 0 hidden)
+ = Active Route, - = Last Active, * = Both

102.102.102.102/32 *[LDP/9] 00:00:02, metric 1
                    > to 192.168.46.4 via em0.0, Push 28

The outcome for traffic forwarding will be the identical to bgp-igp or bgp-igp-both-ribs but not for routing operations

BGP is using table inet.3 to resolve next hops, where as normal IGP routing uses inet.0
Important note is with BGP and inet.3 if inet.0 contains a better route (e.g. better preference) then BGP would use the inet.0 route and traffic would not be forwarded on the LSP causing MPLS VPNs to break.

Other reference

https://www.networkfuntimes.com/moving-lsps-between-inet-3-and-inet-0-on-a-juniper-router/
http://matt.dinham.net/mapping-traffic-lsp-junos-part-1/
https://packetcorner.wordpress.com/2012/10/06/difference-between-traffic-engineering-options-bgp-igp-vs-mpls-forwarding/