Forwarding Table Compression via Prefix Merging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional digital content distribution systems face challenges in efficiently storing routing information due to the increasing number of destination nodes exceeding the capacity of forwarding tables, leading to increased latency and cost prohibitive high-speed memory requirements.
Innovation Solution
A method for compressing forwarding tables by storing network prefixes and associated routing information, where a second network prefix is only written to the forwarding table if it does not overlap with an existing prefix, allowing for the removal of redundant entries and reducing memory requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the number of destination nodes stored in the routing table increases to accommodate the growing Internet, then the routing information becomes more comprehensive and up-to-date, but the number of entries in the forwarding table exceeds its capacity and memory requirements become cost prohibitive
Solution Approach 1:
The patent merges multiple routing table entries that share the same next-hop destination into a single forwarding table entry. By grouping routes with identical next-hop information under a common prefix, the system reduces the total number of forwarding table entries while preserving complete routing information in the routing table. This combining approach allows the forwarding table to scale without proportionally increasing memory capacity requirements.
Solution Approach 2:
The patent segments routing information into two distinct layers: the routing table stores complete, granular routing information for all destination nodes, while the forwarding table stores only the essential next-hop information needed for rapid packet forwarding. This segmentation allows each table to be optimized for its specific function - the routing table for comprehensiveness and the forwarding table for speed and space efficiency.
2Reliability
If the forwarding table stores more routes to cover all destination nodes, then routing coverage is improved, but the memory capacity requirements become cost prohibitive
Solution Approach 1:
The patent combines multiple routing entries that map to the same next-hop into a single forwarding entry using prefix matching. This merging technique ensures that all destination nodes are covered (by storing complete routes in the routing table) while minimizing the number of entries in the memory-constrained forwarding table, thus reducing cost prohibitive memory requirements.
Solution Approach 2:
The patent extracts only the essential next-hop information from the complete routing table entries and stores it in the forwarding table. By taking out only the critical forwarding information and leaving the detailed routing information in the routing table, the system achieves full routing coverage without proportionally increasing forwarding table memory capacity.
3Speed
If the forwarding table is implemented in high-speed memory such as TCAM, then packet forwarding speed is improved, but the cost and complexity increase significantly
Solution Approach 1:
The patent segments the routing information storage function between the routing table (in standard memory) and the forwarding table (in high-speed memory like TCAM). By placing only the essential next-hop information in the high-speed memory and keeping the complete routing information in standard memory, the system achieves fast forwarding speeds without the prohibitive cost and complexity of implementing the entire routing table in high-speed memory.
Solution Approach 2:
The patent merges multiple routing entries into fewer forwarding table entries, reducing the total number of entries that must be stored in expensive high-speed memory. This merging approach maintains fast packet forwarding speeds (by keeping the forwarding table in TCAM) while significantly reducing the memory capacity requirements and associated costs.
4Measurement precision
If the forwarding table stores complete routing information for all destination nodes, then routing accuracy is improved, but the number of entries exceeds forwarding table capacity
Solution Approach 1:
The patent segments routing information into two parts: complete routing information stored in the routing table and condensed next-hop information stored in the forwarding table. This segmentation allows the forwarding table to maintain high routing accuracy for packet forwarding decisions while storing only the essential information needed, thus avoiding entry count overflow.
Solution Approach 2:
The patent merges multiple routing table entries that share the same next-hop into single forwarding table entries using prefix-based matching. This merging preserves routing accuracy by maintaining the complete routing information in the routing table while reducing the forwarding table entry count to fit within capacity constraints.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
One embodiment of the present invention sets forth a technique for compressing a forwarding table. The technique includes storing a first network prefix and a first set of routing information associated with the first network prefix in a routing table, and writing the first network prefix from the routing table to a first entry included in a forwarding table. The technique further includes storing a second network prefix and a second set of routing information associated with the second network prefix in the routing table. The second network prefix is covered by the first network prefix. The technique further includes, based on comparing the first set of routing information to the second set of routing information, determining that the second network prefix should not be written from the routing table to the forwarding table, or removing the second network prefix from the forwarding table.