Multi-destination Traffic Control in Multi-level IP Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional IP network technologies, such as PIM, face challenges in supporting multi-destination traffic forwarding in data center networks, particularly in virtualized environments, leading to inefficient resource utilization and scalability issues with a large number of overlay virtual networks.
Innovation Solution
The implementation of a traffic control method that establishes a bi-directional IP multicast distribution tree across multiple network areas using ISIS and OSPF protocols, with extensions to support inter-area transport, including the use of sub-type-length-value (TLV) for inter-area routers and root address (RTADDR) sub-TLVs, to manage multicast traffic efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If PIM solutions are used for multicast transport, then IP multicast traffic can be carried over IP networks, but network resources are unnecessarily consumed and scalability is limited
Solution Approach 1:
The patent segments the network into multiple areas with border routers that summarize multicast group memberships. This segmentation allows efficient inter-area routing by reducing the amount of multicast routing information that needs to be propagated throughout the entire network, thereby lowering overall network resource consumption while maintaining multicast functionality.
Solution Approach 2:
The patent implements preliminary action by having border routers pre-compute and summarize multicast group membership information before it needs to be routed between areas. This summary mechanism is prepared in advance and reduces the control traffic burden during actual data transmission, addressing the scalability issue.
2Adaptability or versatility
If PIM solutions are applied to DC IP networks with network virtualization overlays, then multicast support can be simulated, but the solution becomes challenging and complex
Solution Approach 1:
The patent introduces border routers as intermediary components between different network areas. These border routers act as mediators that translate and summarize multicast information, enabling overlay network multicast capabilities without requiring full PIM implementation throughout the entire network. This intermediary approach simplifies the overall protocol implementation complexity.
3Ease of operation
If multicast support is implemented in the underlying network to simulate overlay layer 2 broadcast capability, then broadcast functionality is achieved, but control traffic increases and convergence is slower
Solution Approach 1:
The patent implements preliminary action by having border routers pre-compute and advertise multicast summary information in advance. This preliminary summary generation reduces the amount of control traffic needed during actual multicast data transmission and accelerates convergence by having routing information already prepared and cached in the network infrastructure.
Solution Approach 2:
The patent merges multiple multicast group memberships into summarized routing information at border routers. This merging reduces the volume of control traffic that needs to be propagated through the network, thereby reducing control traffic overhead and accelerating routing convergence while maintaining full broadcast capability.
Data Source
AI summary
An internet protocol (IP) traffic control method comprising receiving a message that comprises multicast group information and distribution tree root information from a tree root node in a first network area, determining the border network node is a designated forwarder for the second network area, sending an announcement message that comprises the multicast group information and the distribution tree root information within the second network area, receiving multicast group membership information for a multicast group computing a multicast group membership summary for the first network area and the second network area, sending the multicast group membership summary to first network area and the second network area, computing a distribution tree using the multicast group information and the distribution tree root information, pruning the distribution tree to generate a multicast tree, and communicating multicast data traffic between the first network area and the second network area along the multicast tree.