Ethernet Forwarding Database Hash Table Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Ethernet bridge devices face inefficiencies in data packet routing due to hash conflicts and limited capacity in hash tables, leading to increased flooding and lookup failures as the table becomes full, especially when MAC addresses are not optimally distributed across hash entries.
Innovation Solution
Implementing a MAC address table with multiple hash table columns, each with an independent hash function and usage count, along with a conflict resolution mechanism that relocates entries to free slots based on usage counts, allowing for concurrent access and increased table utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single hash table is used for forwarding database storage, then the structure is simple, but hash conflicts increase and table utilization decreases as the table becomes full
Solution Approach 1:
The forwarding database hash table is divided into multiple columns (e.g., 2 columns), each with its own independent hash function. This segmentation allows entries to be distributed across multiple columns, reducing hash conflicts within each column and improving overall table utilization and lookup efficiency.
2Quantity of substance
If hash table capacity is increased to accommodate more MAC addresses, then more entries can be stored, but hash conflicts increase and lookup efficiency decreases
Solution Approach 1:
By segmenting the hash table into multiple columns with independent hash functions, the system can store more MAC address entries while maintaining lookup efficiency. Each column handles a portion of the hash space, distributing the load and reducing collision frequency even as total capacity increases.
Solution Approach 2:
The system changes the hash function parameter by using different independent hash functions for each column. This allows the same key (MAC address) to be mapped to different positions in different columns, enabling better distribution of entries and maintaining efficiency at higher capacities.
3Ease of manufacture
If a single hash function is used for all entries, then the implementation is simple, but non-optimal distribution of MAC addresses leads to hash conflicts and flooding
Solution Approach 1:
The system segments the hash function application by assigning different independent hash functions to different columns. While this increases implementation complexity slightly, it dramatically improves entry distribution quality by reducing the probability of hash conflicts and eliminating the need for flooding.
4Quantity of substance
If the forwarding database table is filled to maximum capacity, then storage utilization is maximized, but new entries cause conflicts and require flooding
Solution Approach 1:
Segmenting the table into multiple columns with independent hash functions allows new entries to be inserted more easily even when the table is full. Each column maintains its own hash space, so there are more available positions across the entire structure, reducing the need for flooding and improving insertion ease.
Data Source
AI summary
A technique is described for retrieving data from a hash table in an Ethernet bridge. The table has a plurality of columns, each of which has a hash function, thereby creating respective different hash key values.