Dynamic Page Size Recommendations for Cloud Multi-Tenant Fairness

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In multi-tenant cloud computing environments, unbalanced data load generation from certain subscribers can lead to overloading, data starvation, and compromised performance, as existing technologies often rely on fixed page sizes and request counts that fail to account for real-time platform and subscription plan variations.

Innovation Solution

A system that generates dynamic page size and sequential request count recommendations based on real-time cloud platform and external database statuses, as well as subscription plans, using a combination of machine learning and ruleset-based approaches to ensure fair resource distribution across tenants.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If fixed page sizes and request counts are used, then system simplicity is maintained, but resource distribution fairness deteriorates and performance is compromised

Engineering Contradiction:
Improvesystem simplicityVSAvoidresource distribution fairness
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent implements dynamic page size and sequential request count recommendations that adapt to real-time platform status and subscription plans. The system transitions from fixed static values to dynamically adjusted values based on current system conditions, tenant priorities, and resource availability, thereby achieving fair resource distribution while maintaining acceptable system complexity through automated machine learning-based adjustments.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the parameters of page size and request count from fixed constants to variable values determined by machine learning models. These parameters are continuously adjusted based on platform status, subscription plans, and performance data, enabling the system to optimize resource allocation fairness without requiring complex manual configuration.

Inventive Principle:
Principle #35Parameter changes

2Reliability

If dynamic recommendations based on real-time status are implemented, then resource distribution fairness is improved, but system complexity increases

Engineering Contradiction:
Improveresource distribution fairnessVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements a self-service system where machine learning models automatically generate page size and request count recommendations based on real-time platform status and subscription plans. The system self-adjusts without manual intervention, using automated feedback loops that monitor performance and continuously refine recommendations, thereby reducing the operational complexity burden despite the increased system intelligence.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent incorporates feedback mechanisms where performance data from the cloud platform is continuously collected and used to train and refine machine learning models. This feedback loop enables the system to automatically improve its resource allocation decisions over time, balancing the increased complexity of dynamic adjustments with the benefit of self-optimizing fairness.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If machine learning engine is trained with performance data, then recommendation accuracy is improved, but processing time and computational resources increase

Engineering Contradiction:
Improverecommendation accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary action by training the machine learning engine with historical performance data before it is needed for real-time recommendations. This offline training phase prepares the model to make rapid predictions during operation, separating the computationally intensive training process from the time-sensitive recommendation generation process, thereby minimizing processing time delays in production.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4538883A1Dynamic learning based resource utilizing fairness in a cloud multi-tenant environment
Publication Date: 2025.04.16 SAP SE
  • EP4538883A1 patent drawingFigure 1
  • EP4538883A1 patent drawingFigure 2
  • EP4538883A1 patent drawingFigure 3

AI summary

A system for executing a multi-tenant application includes at least one processor and at least one memory storing program instructions. The multi-tenant application generates one or more page size recommendations and one or more sequential request count recommendations for one or more calls to an external database. The multi-tenant application performs a first call to the external database using a page size which is based on a first page size recommendation, where the page size specifies a number of records to retrieve from the external database. The multi-tenant application also performs, in a sequential manner by the multi-tenant application, the first call and a number of subsequent calls to the external database, where the number of subsequent calls is based on a first sequential request count recommendation. Related methods and computer program products are also provided.