- Drivers for Scalable Next-Generation Multicast VPN
- Juniper Networks
Several emerging multicast applications require virtualization to support service delivery, traffic separation, and new business models. A more scalable and flexible approach to implementing multicast VPN (MVPN) is being deployed in service provider, financial, and enterprise networks to effectively support these emerging applications, while achieving greater economies of scale.
The volume of multicast traffic has been growing primarily based on the emergence of video-based applications. There are a growing number of Layer 3 VPN customers who have IP multicast traffic. (These Layer 3 VPNs are commonly known as 2547 VPNs and are based on the original RFC 2547.) Service providers who have an installed base of Layer 3 VPN for unicast service are looking at upselling with new media-rich solutions to increase ARPU while achieving operational efficiency. The significant interest in IPTV services and wholesale business models are driving the need to consider more scalable ways to deliver multicast services. Similarly, the delivery of multicast-dependent financial services requires scalable and reliable MVPN infrastructures. In addition to Layer 3 MPLS VPN services, service providers and cable operators are beginning to offer virtual private LAN service (VPLS) to small and medium businesses, as well as to large enterprises. There is also a growing momentum towards IPv6-enabled services based on new mobile applications. This diversity and breadth of services pose a challenge for operators to create an infrastructure that supports Layer 2 VPNs, Layer 3 VPNs, IPv4, IPv6, unicast, and multicast traffic. The difficulty is particularly true for virtual services that require complex control and data plane operations. Another challenge is to support emerging multicast applications incrementally on top of existing Layer 3 VPN and VPLS infrastructures without adding operational complexity.
Download Paper
(Webtorials registration required for downloads. Click here if you forgot your username/password.)
The volume of multicast traffic has been growing primarily based on the emergence of video-based applications. There are a growing number of Layer 3 VPN customers who have IP multicast traffic. (These Layer 3 VPNs are commonly known as 2547 VPNs and are based on the original RFC 2547.) Service providers who have an installed base of Layer 3 VPN for unicast service are looking at upselling with new media-rich solutions to increase ARPU while achieving operational efficiency. The significant interest in IPTV services and wholesale business models are driving the need to consider more scalable ways to deliver multicast services. Similarly, the delivery of multicast-dependent financial services requires scalable and reliable MVPN infrastructures. In addition to Layer 3 MPLS VPN services, service providers and cable operators are beginning to offer virtual private LAN service (VPLS) to small and medium businesses, as well as to large enterprises. There is also a growing momentum towards IPv6-enabled services based on new mobile applications. This diversity and breadth of services pose a challenge for operators to create an infrastructure that supports Layer 2 VPNs, Layer 3 VPNs, IPv4, IPv6, unicast, and multicast traffic. The difficulty is particularly true for virtual services that require complex control and data plane operations. Another challenge is to support emerging multicast applications incrementally on top of existing Layer 3 VPN and VPLS infrastructures without adding operational complexity.
Download Paper
(Webtorials registration required for downloads. Click here if you forgot your username/password.)
As applications like videocasting pervade both the corporate and the consumer worlds, mullticasting is gaining tremendously in importance. Though the application is far removed from many of the original uses of "multicast," it has resurfaced as a major issue. In particular, not that multiple high-definition video streams of the same information need to be delivered, controlling the number of "copies" of the same information on the network is of utmost importance.
This paper does a great job of describing the issue, and, as something that will be greatly appreciated by our community, provides proposed solutions in a good deal of technical detail.