Dual-Table Packet Lookup for TOR Switch Power Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Top-of-rack (TOR) switches face challenges due to the high power and area consumption of ternary content-addressable memory (TCAM) used for packet lookups, limiting flexibility and increasing costs, as well as inefficiencies in conventional systems that rely heavily on TCAMs.
Innovation Solution
A dual-table system is implemented, with a primary binary table for common searches and a secondary ternary table for less frequent searches, utilizing hash logic to perform binary and ternary searches respectively, and dynamic memory to manage rule updates, reducing power consumption and increasing flexibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If TCAM is used for packet lookups in TOR switches, then lookup speed is improved, but power consumption and area increase significantly
Solution Approach 1:
The patent divides the single TCAM into multiple segments (first TCAM, second TCAM, third TCAM) with different capacities and power characteristics. The system segments packet lookup operations into multiple stages, using smaller TCAM segments for common lookups and larger segments only when necessary, thereby reducing overall power consumption while maintaining lookup speed.
Solution Approach 2:
Different TCAM segments are assigned different qualities based on their capacity and power consumption characteristics. The first TCAM is optimized for speed with lower power, the second for moderate capacity, and the third for high capacity but only activated when needed. This local differentiation allows the system to match resource quality to operational needs.
2Speed
If TCAM is used for packet lookups in TOR switches, then lookup speed is improved, but area consumption increases
Solution Approach 1:
The patent segments the TCAM resource into multiple smaller units distributed across different memory modules. Instead of one large TCAM occupying significant area, the system uses multiple smaller TCAMs that can be distributed and activated selectively, reducing the area footprint while maintaining lookup performance through parallel or sequential access.
Solution Approach 2:
The system dynamically activates only the necessary TCAM segments based on current operational requirements. Not all TCAM segments are active simultaneously, allowing the physical area to be effectively reduced by deactivating unused segments and putting them into low-power or standby states.
3Device complexity
If conventional single-table TCAM system is used, then implementation is simple, but flexibility and adaptability are limited
Solution Approach 1:
The patent divides the rule set into multiple segments stored in different TCAMs, allowing independent configuration and management of each segment. This segmentation enables flexible adaptation to different network requirements by activating specific segments as needed, while the modular structure maintains relative implementation simplicity through standardized interfaces.
Solution Approach 2:
The system implements dynamic rule installation and activation across multiple TCAM segments. Rules can be added, removed, or modified in specific segments without affecting the entire system, providing flexibility and adaptability while maintaining manageable complexity through localized changes.
Data Source
AI summary
A method for performing packet lookups is provided. Packets (which each have a body and a header) are received and parsed to parsing headers. A hash function is applied to each header, and each hashed header is compared with a plurality of binary rules stored within a primary table, where each binary rule is a binary version of at least one ternary rule from a first set of ternary rules. For each match failure with the plurality of rules, a secondary table is searched using the header associated with each match failure, where the secondary table includes a second set of ternary rules.