ADC Cluster Hash Table for Adaptive Traffic Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ADC clustering techniques face limitations such as requiring all devices to have the same capacity, inflexible configuration changes, and physical co-location of switches, which hinder dynamic capacity adjustments and elasticity in cloud-computing environments, leading to inefficiencies in high availability and adaptability.
Innovation Solution
A method and system for managing an ADC cluster using a hash table with multiple buckets, where each active ADC is allocated a bucket, and network elements are instructed to distribute traffic based on the hash table, allowing for dynamic allocation and reconfiguration to match changing load conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If ADC devices are clustered to increase capacity, then the system can handle more traffic, but all devices must have the same capacity which reduces flexibility
Solution Approach 1:
The patent segments the ADC cluster into multiple groups based on device capacity characteristics. Instead of requiring all devices to have identical capacity, the system divides devices into segments (groups) where each group contains devices with similar capacity levels. This allows the cluster to accommodate heterogeneous device capacities while maintaining efficient load distribution within each segment.
2Adaptability or versatility
If configuration changes are made to ADC cluster, then adaptability improves, but disruptions to traffic distribution occur
Solution Approach 1:
The patent implements preliminary action by pre-calculating and preparing hash table configurations before actual ADC devices are added or removed from the cluster. When configuration changes are needed, the system has already prepared the appropriate hash table structures, allowing for seamless transitions without disrupting ongoing traffic distribution. This pre-computation approach ensures that adaptability changes can be applied without causing service interruptions.
3Speed
If switches are physically co-located with ADC devices, then network performance improves, but system flexibility and scalability are reduced
Solution Approach 1:
The patent makes the hash table structure universal by designing it to work independently of physical device locations. The hash table can be implemented in various network elements (switches, routers, or centralized controllers) without requiring specific physical co-location. This universal design allows the system to maintain high performance while providing flexibility in deployment architectures, enabling both co-located and distributed switch configurations.
4Measurement precision
If hash table is updated frequently to reflect capacity changes, then traffic distribution accuracy improves, but system overhead increases
Solution Approach 1:
The patent applies preliminary action by pre-computing hash table configurations before ADC devices are added or removed. Instead of updating the hash table in real-time as devices change, the system prepares the appropriate hash table structure in advance. This approach maintains high traffic distribution accuracy while avoiding the overhead of frequent real-time updates, as the pre-computed tables are ready for immediate deployment when configuration changes occur.
Data Source
AI summary
A system and method for managing an application delivery controller (ADC) cluster including a plurality of ADCs are provided. The method includes creating a hash table including a plurality of buckets, wherein a number of the plurality of buckets is a multiple of a maximum number of active ADCs that can be supported by the ADC cluster; allocating, to each active ADC of the ADC cluster, one of the plurality of buckets; and instructing at least one network element to distribute traffic to and from the active ADCs based on the hash table.