- play_arrow Overview
- play_arrow Managing Group Membership
- play_arrow Configuring IGMP and MLD
- play_arrow Configuring IGMP Snooping
- IGMP Snooping Overview
- Overview of Multicast Forwarding with IGMP Snooping or MLD Snooping in an EVPN-VXLAN Environment
- Configuring IGMP Snooping on Switches
- Example: Configuring IGMP Snooping on Switches
- Example: Configuring IGMP Snooping on EX Series Switches
- Verifying IGMP Snooping on EX Series Switches
- Changing the IGMP Snooping Group Timeout Value on Switches
- Monitoring IGMP Snooping
- Example: Configuring IGMP Snooping
- Example: Configuring IGMP Snooping on SRX Series Devices
- Configuring Point-to-Multipoint LSP with IGMP Snooping
- play_arrow Configuring MLD Snooping
- Understanding MLD Snooping
- Configuring MLD Snooping on an EX Series Switch VLAN (CLI Procedure)
- Configuring MLD Snooping on a Switch VLAN with ELS Support (CLI Procedure)
- Example: Configuring MLD Snooping on EX Series Switches
- Example: Configuring MLD Snooping on SRX Series Devices
- Configuring MLD Snooping Tracing Operations on EX Series Switches (CLI Procedure)
- Configuring MLD Snooping Tracing Operations on EX Series Switch VLANs (CLI Procedure)
- Example: Configuring MLD Snooping on EX Series Switches
- Example: Configuring MLD Snooping on Switches with ELS Support
- Verifying MLD Snooping on EX Series Switches (CLI Procedure)
- Verifying MLD Snooping on Switches
- play_arrow Configuring Multicast VLAN Registration
-
- play_arrow Configuring Protocol Independent Multicast
- play_arrow Understanding PIM
- play_arrow Configuring PIM Basics
- Configuring Different PIM Modes
- Configuring Multiple Instances of PIM
- Changing the PIM Version
- Optimizing the Number of Multicast Flows on QFabric Systems
- Modifying the PIM Hello Interval
- Preserving Multicast Performance by Disabling Response to the ping Utility
- Configuring PIM Trace Options
- Configuring BFD for PIM
- Configuring BFD Authentication for PIM
- play_arrow Routing Content to Densely Clustered Receivers with PIM Dense Mode
- play_arrow Routing Content to Larger, Sparser Groups with PIM Sparse Mode
- Understanding PIM Sparse Mode
- Examples: Configuring PIM Sparse Mode
- Configuring Static RP
- Example: Configuring Anycast RP
- Configuring PIM Bootstrap Router
- Understanding PIM Auto-RP
- Configuring All PIM Anycast Non-RP Routers
- Configuring a PIM Anycast RP Router with MSDP
- Configuring Embedded RP
- Configuring PIM Filtering
- Examples: Configuring PIM RPT and SPT Cutover
- Disabling PIM
- play_arrow Configuring Designated Routers
- play_arrow Receiving Content Directly from the Source with SSM
- Understanding PIM Source-Specific Mode
- Example: Configuring Source-Specific Multicast
- Example: Configuring PIM SSM on a Network
- Example: Configuring an SSM-Only Domain
- Example: Configuring SSM Mapping
- Example: Configuring Source-Specific Multicast Groups with Any-Source Override
- Example: Configuring SSM Maps for Different Groups to Different Sources
- play_arrow Minimizing Routing State Information with Bidirectional PIM
- play_arrow Rapidly Detecting Communication Failures with PIM and the BFD Protocol
- play_arrow Configuring PIM Options
- play_arrow Verifying PIM Configurations
-
- play_arrow Configuring Multicast Routing Protocols
- play_arrow Connecting Routing Domains Using MSDP
- play_arrow Handling Session Announcements with SAP and SDP
- play_arrow Facilitating Multicast Delivery Across Unicast-Only Networks with AMT
- play_arrow Routing Content to Densely Clustered Receivers with DVMRP
-
- play_arrow General Multicast Options
- play_arrow Bit Index Explicit Replication (BIER)
- play_arrow Prevent Routing Loops with Reverse Path Forwarding
- play_arrow Use Multicast-Only Fast Reroute (MoFRR) to Minimize Packet Loss During Link Failures
- play_arrow Enable Multicast Between Layer 2 and Layer 3 Devices Using Snooping
- play_arrow Configure Multicast Routing Options
- play_arrow Controller-Based BGP Multicast Signaling
-
- play_arrow Troubleshooting
- play_arrow Knowledge Base
-
- play_arrow Configuration Statements and Operational Commands
BGP-MVPN Inter-AS Option B Overview
This topic provides an overview of Junos support for Inter-Autonomous System (AS) Option B, which is achieved by extending Border Gateway Protocol Multicast Virtual Private Network (BGP-MVPN) to support Inter-AS scenarios using segmented provider tunnels (p-tunnels). Junos OS also support Option A and Option C unicast with non-segmented p-tunnels, support for which was introduced in Junos OS 12.1. See the links below for more information on these options.
Inter-AS support for multicast traffic is required when an L3VPN results in two or more ASes that are using BGP-MVPN. The ASes may be administered by the same authority or by different authorities. When using BGP-MVPN Inter-AS Option B with segmented p-tunnels, the p-tunnel segmentation is performed at the Autonomous System Border Router (ASBRs). The ASBRs also perform BGP-MVPN signaling and form the data plane.
Setting up Inter-AS Option B with segmented p-tunnels can be complex, but the configuration does provide the following advantages:
Independence. Different administrative authorities can choose whether or not to allow topology discovery of their AS by the other ASes. That is, each AS can be separately controlled by a different independent authority.
Heterogeneity. Different p-tunnel technologies can be used within a given AS (as might be the case when working with heterogeneous networks that now must be combined).
Scale. Inter-AS Option B with segmented p-tunnels avoids the potential for ASBR bottleneck that can happen when Intra-AS p-tunnels are set up across ASes using non-segmented p-tunnels. (Unicast branch LSPs with inclusive p-tunnels can all have to transit through the ASBRs. In this case, for IR, the pinch point becomes data-plane scale. For RSVP-TE it becomes P2MP control-plane scale, due to the high number of RSVP refresh messages passing through the ASBRs).
The supported Junos implementation of Option B uses RSVP-TE p-tunnels for all segments, and MVPN Inter-AS signaling procedures. Multicast traffic is forwarded across AS boundaries over a single-hop labeled LSP. Inter-AS p-tunnels have two segments: an ASBR-ASBR segment, called Inter-AS segment and the ASBR-PE segment called Intra-AS segment. (Static RSVP-TE, IR , PIM-ASM, and PIM-SSM p-tunnels are not supported.)
MVPN Intra-AS AD routes are not propagated across the AS boundary. The Intra-AS inclusive p-tunnels advertised in Type-1 routes are terminated at the ASBRs within each AS. Route learning for both unicast and multicast traffic occurs only through Option B.
The ASBR originates an Inter-AS AD (Type-2) route into eBGP, which may include tunnel attributes for an Inter-AS p-tunnel (called an Inter-AS, or ASBR-ASBR p-tunnel segment). The Type-2 route contains the ASBR's route distinguisher (RD), which is unique per VPN and per ASBR, and its AS number. The tunnel is set up between two directly connected ASBRs in neighboring ASes, and it is always a single-hop point-topoint (P2P) LSP.
An ASBR in the originating AS forwards all multicast traffic received over the inclusive p-tunnel into the Inter-AS p-tunnel. An ASBR in the adjacent AS propagates the received Inter-AS route into its own AS over iBGP, but only after rewriting the Provider Multicast Service Interface (PMSI) tunnel attributes and modifying the next-hop of the Multiprotocol Reachable (MP_REACH_NRLI) attribute with a reachable address of the ASBR (next-hop self rewrite). When an ASBR propagates the Type-2 route over iBGP, it can choose any p-tunnel type supported within its AS, although the supported Junos implementation of Option B uses RSVP-TE p-tunnels only for all segments.
At the ASBRs, traffic received over the upstream p-tunnel segment is forwarded over the downstream p-tunnel segment. This process is repeated at each AS boundary. The resulting Inter-AS p-tunnel is comprised of alternating Inter-AS and Intra-AS p-tunnel segments (thus the name, “segmented p-tunnel”).
Option B with segmented p-tunnels is not without drawbacks.:
The ASBRs distribute both VPN routes and routes in the master instance. They may thus become a bottleneck.
With a large number of VPNs, the ASBR can run out of labels because each unicast VPN route requires one.
Per VPN packet flow accounting cannot be performed at the ASBR.
Unless route-targets are rewritten at the AS boundaries, the different service providers must agree on VPN route-targets (this is that same as for option-C)
The ASBRs must be capable of MVPN signaling and support Inter-AS MVPN procedures.