Network Agent Throttling for Server Overload Prevention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large-scale networked enterprise computing systems face challenges in managing system management messages due to server overload and agent starvation, as hundreds or thousands of agents sending data simultaneously can flood the server or prevent others from communicating effectively.
Innovation Solution
Implementing a throttling mechanism where agents limit the number of messages sent to the server within predetermined time periods, using both send throttling and queue throttling techniques, to prevent server overload and manage resource usage, allowing for a controlled flow of messages and reducing CPU usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If all agents send data to the server simultaneously, then complete data collection is achieved, but the server becomes flooded and cannot handle messages
Solution Approach 1:
The patent implements periodic throttling where agents are assigned time slots and can only send messages during their allocated period. This periodic action distributes the message load across time, preventing server flooding while ensuring all agents eventually get to send their data.
Solution Approach 2:
The patent segments the message sending process by dividing agents into groups and assigning them different time slots. This segmentation allows the server to handle messages from one group at a time, preventing overload while maintaining complete data collection from all agents.
2Productivity
If agents send messages continuously, then data transmission speed is maximized, but other agents starve and cannot send messages
Solution Approach 1:
The patent uses periodic time slots to ensure fair message transmission. Each agent gets a dedicated period during which they can send messages without interruption, preventing starvation while maintaining efficient data collection through continuous cycling through all agents.
Solution Approach 2:
The server tracks which agents have sent messages during each time slot and adjusts the scheduling accordingly. This feedback mechanism ensures that agents who haven't sent messages yet get their turn, preventing starvation while maximizing overall transmission efficiency.
3Reliability
If a cluster of redundant servers is built, then server overload is prevented, but system complexity increases
Solution Approach 1:
The patent implements dynamic message throttling at the agent level, where the sending behavior is adjusted based on current system conditions. This dynamic approach prevents server overload without requiring additional servers, maintaining system simplicity while improving reliability.
Solution Approach 2:
The agents themselves implement throttling logic to control their own message sending behavior. This self-service approach distributes the load management function to the agents rather than requiring a complex server architecture, preventing overload while keeping the system simple.
Data Source
AI summary
Method and apparatus for processing system management messages are described herein. In one embodiment, a throttling policy associated with a network node is determined for throttling system management messages to be sent to a system management server over a network. The system management messages are then sent to the system management server according to the throttling policy, where the system management server is configured to manage multiple network nodes over the network. Other methods and apparatuses are also described.