Prompt Intensity Control for Distributed Server Load Timing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional server processes experience increased load at specific times, leading to concentrated timing of processing and communication, which can result in server overload.
Innovation Solution
A control method that generates a pattern extended or compressed along a time axis for prompting users to distribute the timing of their operations, reducing the instantaneous increase in server load by adjusting the start time and intensity of prompts based on probable operation start times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the server executes processes at predetermined time slots (other than hour beginning), then the execution timing is standardized and manageable, but the server load increases at concentrated timing
Solution Approach 1:
The patent applies periodic action by using time slot management to schedule process executions at regular intervals. The server divides time into discrete slots and executes processes periodically within these slots, avoiding continuous execution while maintaining standardized timing management. This resolves the contradiction by providing structured timing control without causing load concentration at specific moments.
Solution Approach 2:
The patent employs dynamics by adjusting the execution timing within allowed time slots based on current server load conditions. Rather than executing processes at fixed times regardless of conditions, the system dynamically selects appropriate time slots for execution, distributing the load across different periods while maintaining standardized slot-based management.
2Quantity of substance
If the prompt start time is distributed to reduce server load, then the processing timing is distributed, but the prompt may start after the desired prompt start time causing user perception of system failure
Solution Approach 1:
The patent applies preliminary action by preparing and buffering the prompt message in advance before the desired start time. The system generates the prompt content and readies the transmission beforehand, then releases it at the optimal moment. This ensures the prompt appears timely from the user perspective while the actual processing can be distributed to avoid server load concentration.
Solution Approach 2:
The patent uses an intermediary mechanism (prompt buffer/cache) that decouples the prompt generation from the actual server processing. The prompt is prepared as an intermediate representation that can be released independently of the heavy server operations, allowing timing distribution for load reduction without affecting the perceived timing accuracy for users.
3Ease of operation
If multiple users set the same operation start time, then the scheduling is simple for each user, but the server load increases significantly at that time
Solution Approach 1:
The patent applies parameter changes by introducing a time slot parameter that modifies the execution timing without changing the user's intended operation start time. The system accepts simple user schedules, then transforms them into distributed execution schedules by applying time slot offsets. This maintains ease of operation for users while distributing the actual server load across different time parameters.
Solution Approach 2:
The patent segments the集中 execution time into multiple time slots, distributing the same operation across different temporal segments. When multiple users set the same operation start time, the system divides their executions into different time slot segments, reducing peak load while maintaining the simplicity of user schedule setting.
Data Source
AI summary
A time that is a predetermined period before an operation start time which is a time in a schedule at which a first device starts an operation is obtained as a prompt start time, a pattern extended or compressed along a time axis from an initial pattern indicating a temporal change in an intensity of a prompt is generated; and a second device is caused to start outputting the prompt at the prompt start time and to output the prompt with an intensity indicated by the pattern generated.