Decentralized Scheduling for Network Test Agents
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication network testing methods face challenges with the 'Thundering Herd' problem, where multiple test agents simultaneously running tests overload the system, skewing results and requiring complex coordination to distribute the testing load effectively.
Innovation Solution
A software-driven mechanism for deterministic decentralized scheduling of test agents, using unique identifiers for each agent and test, allows them to determine optimal test times independently, eliminating the need for centralized coordination and preventing system overload.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple test agents simultaneously execute tests from many test locations, then testing coverage and productivity are improved, but the system becomes overloaded and test results are skewed
Solution Approach 1:
The patent segments the testing timeline into discrete time slots and assigns each test agent a specific time window for execution. This temporal segmentation allows multiple agents to operate simultaneously without interfering with each other, maintaining both productivity and result accuracy by preventing the 'thundering herd' effect where all agents execute at once.
Solution Approach 2:
The system incorporates feedback mechanisms where test agents report their execution status and the system adjusts scheduling accordingly. This feedback loop enables dynamic coordination among agents, allowing the system to optimize testing coverage while preventing overload by redistributing agents to different time windows based on real-time system state.
2Productivity
If centralized coordination is used to schedule test agents, then testing load is distributed effectively, but system complexity and coordination overhead increase
Solution Approach 1:
Each test agent independently calculates its own optimal execution time based on locally available information such as its unique identifier and the current system state. This self-service approach eliminates the need for complex centralized scheduling while achieving effective load distribution, as each agent autonomously determines when to execute without requiring continuous coordination with a central authority.
Solution Approach 2:
The system performs preliminary scheduling actions by pre-calculating time windows and assigning them to agents before execution begins. This preliminary action simplifies the overall coordination mechanism by establishing the schedule in advance, reducing the complexity of real-time coordination while maintaining effective load distribution across the testing system.
3Reliability
If test agents communicate extensively to coordinate scheduling, then testing impact on the system is minimized, but communication overhead and time to execute tests increase
Solution Approach 1:
The patent applies partial action by having agents communicate only the minimum necessary information to determine their execution time windows, rather than extensive coordination. Each agent shares only essential parameters (such as unique identifiers and basic system state) sufficient for scheduling decisions, minimizing communication overhead while still achieving effective load distribution and reducing testing impact on the system.
Data Source
AI summary
A method and system for testing of a communication network utilizing a plurality of test agents in the communication network using a deterministic decentralized scheduling of the plurality of test agents to test a single network service. Determining an optimal time period for executing a test of the single network service by each of the plurality of test agents without requiring global knowledge of the system and without requiring centralized coordination of individual agent test schedules.