Cloud Server Time-Slot Scheduling to Prevent TOR Request Collisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In cloud service systems, data transmission between servers is affected by request collisions at TOR switches when multiple servers send requests simultaneously, leading to potential losses without a buffer, impacting stability.
Innovation Solution
Implement a method where servers in a group are assigned unique identifiers, allowing them to send requests at distinct time domain locations within a slot, ensuring no collisions and using time adjustments based on transmission durations to synchronize with the arbiter.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If multiple servers send requests simultaneously through the same TOR switch without a buffer, then the system structure remains simple, but request collisions occur causing data transmission instability
Solution Approach 1:
The patent divides time into periodic slots and assigns different servers to transmit requests at different time domain locations within each slot. This periodic time-division multiplexing approach allows multiple servers to access the TOR switch without collisions while maintaining a simple switch structure without buffers.
Solution Approach 2:
The patent segments the time domain into multiple distinct locations within each slot, assigning each server a unique time domain location. This temporal segmentation prevents request collisions by ensuring that only one server transmits at any given time domain location, thereby improving reliability without adding buffer memory.
2Reliability
If time domain locations are assigned to different servers, then request collisions are prevented improving reliability, but time synchronization complexity increases
Solution Approach 1:
Each server autonomously determines its own time domain location based on its identifier and the predefined slot structure. The arbiter also independently calculates time domain locations for response packets. This self-service approach minimizes synchronization complexity by eliminating the need for complex centralized coordination mechanisms.
Solution Approach 2:
The patent uses server identifiers as parameters to determine time domain locations through a predefined mapping relationship. By changing the identifier parameter, each server automatically obtains its unique time domain location, simplifying the synchronization mechanism to a straightforward parameter-based assignment rather than complex real-time coordination.
Data Source
AI summary
An information transmission method and a related device applied to a cloud service system are provided. The cloud service system includes an arbiter, a first switch, and a first server group. The arbiter is connected to the first server group by using the first switch. The method includes: a first server obtains a first identifier of the first server in the first server group. The first server determines, in a first slot, a first time domain location corresponding to the first identifier. The first time domain location is a time domain location at which the first switch receives a first request to be processed by the arbiter, and different identifiers correspond to different time domain locations in the first slot. The first server sends the first request to the first switch based on the first time domain location.