Software-Defined Network Flow Entry Lookup via Constant-Size Keys

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Software-defined networks using OPENFLOW-enabled switches face performance degradation due to the need to sift through linked list data structures for flow entries, leading to scalability issues as the number of entries increases.

Innovation Solution

Implementing a system with databases configured to store flow entries using keys of constant size, allowing for efficient searching and operations like lookup, addition, and deletion independent of the number of flow entries, utilizing trie data structures and hash tables to maintain consistent performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If OPENFLOW-enabled switches maintain flow entries as linked list data structures, then the system is simple to implement, but performance degrades significantly as the number of flow entries increases

Engineering Contradiction:
Improveease of implementationVSAvoidperformance
Core Design Contradiction:
Ease of manufactureVSProductivity

Solution Approach 1:

The patent changes the fundamental parameter of data structure organization from linked lists to hash tables. This transformation changes the time complexity of flow entry operations from O(n) linear search to O(1) constant-time operations, directly resolving the performance degradation issue while maintaining implementation feasibility through well-established hash table algorithms

Inventive Principle:
Principle #35Parameter changes

2Quantity of substance

If the number of flow entries increases to handle more network traffic, then network capacity increases, but search time increases linearly

Engineering Contradiction:
Improvenumber of flow entriesVSAvoidsearch time
Core Design Contradiction:
Quantity of substanceVSLoss of time

Solution Approach 1:

The patent transforms the search mechanism by changing from sequential traversal of linked list nodes to direct indexing via hash table buckets. The search time parameter becomes independent of the number of flow entries, achieving O(1) complexity where search time remains constant regardless of how many flow entries are stored

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces hash functions as an intermediary mechanism between the flow entry key and its storage location. This intermediary transforms the direct search problem into a two-step process: hash computation followed by direct access, eliminating the need to traverse through all flow entries and achieving constant-time search performance

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP2822240B1Systems and methods for increasing the scalability of software-defined networks
Publication Date: 2016.08.17 JUNIPER NETWORKS INC
  • EP2822240B1 patent drawingFigure 1
  • EP2822240B1 patent drawingFigure 2
  • EP2822240B1 patent drawingFigure 3

AI summary

A computer-implemented method for increasing the scalability of software-defined networks may include (1) maintaining a set of databases collectively configured to (i) store a set of flow entries that direct network traffic within a software-defined network and (ii) facilitate searching the set of flow entries based at least in part on at least one key whose size remains substantially constant irrespective of the number of flow entries within the set of flow entries, (2) detecting a request to perform an operation in connection with a flow of data packets within the software-defined network, (3) identifying at least one attribute of the flow of data packets in the request, and then (4) searching, using the attribute of the flow of data packets as a database key, at least one database within the set of databases to facilitate performing the operation. Various other methods, systems, and apparatuses are also disclosed.