Backup Orchestrator Using Prediction Models for Resource Availability

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing backup systems generate backups at scheduled times, which can lead to computational overload and negatively impact the ability of devices to provide other services, resulting in phantom slowdowns and inefficient use of computing resources.

Innovation Solution

A system that uses a backup manager to predict future computing resource availability based on past behaviors, allowing for the generation of unscheduled backups at times when sufficient resources are available, thereby reducing the impact on other services and improving user experience.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If backups are generated at scheduled times, then backup operations can be performed regularly, but computational overload occurs and other services are negatively impacted

Engineering Contradiction:

Improvebackup reliabilityVSAvoidcomputational overload

Core Design Contradiction:

ReliabilityVSObject-generated harmful factors

Solution Approach 1:

The backup system transitions from static scheduled timing to dynamic prediction-based timing. The backup orchestrator continuously monitors computing resource availability and uses prediction models to determine optimal backup moments, allowing the system to adapt to changing resource conditions rather than following rigid schedules

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system implements feedback loops where the backup orchestrator monitors actual computing resource availability, compares it with predictions, and continuously refines the prediction model. This feedback mechanism enables the system to learn from past performance and improve future backup scheduling decisions

Inventive Principle:

Principle #23Feedback

2Stability of the object's composition

If backups are generated at scheduled times, then backup operations are consistent, but phantom slowdowns occur and user experience deteriorates

Engineering Contradiction:

Improvebackup consistencyVSAvoiduser experience

Core Design Contradiction:

Stability of the object's compositionVSEase of operation

Solution Approach 1:

The system performs preliminary actions by continuously monitoring and predicting future computing resource availability before backup events occur. The prediction model prepares advance insights about resource conditions, enabling the system to proactively schedule backups during optimal windows before resource constraints arise

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The backup system serves itself by autonomously determining optimal backup timing based on its own monitoring of computing resource availability. The prediction model and orchestrator work independently to self-regulate backup scheduling without external intervention, adapting to the system's own resource patterns

Inventive Principle:

Principle #25Self-service

3Productivity

If backups are generated at scheduled times, then backup frequency is maintained, but computing resources are inefficiently utilized

Engineering Contradiction:

Improvebackup frequencyVSAvoidcomputing resource efficiency

Core Design Contradiction:

ProductivityVSUse of energy by moving object

Solution Approach 1:

The system changes the timing parameter of backup operations from fixed scheduled values to dynamically predicted values based on computing resource availability. This parameter transformation allows backups to occur at moments when resources are naturally abundant, converting what would be wasted resource capacity into productive backup operations

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11467915B2System and method for backup scheduling using prediction models

Publication Date: 2022.10.11 EMC IP HLDG CO LLC

AI summary

A backup manager for providing backup services includes storage and a backup orchestrator. The persistent storage stores prediction models. The backup orchestrator obtains a computing resource availability for a client; generates a prediction model of the prediction models using the computing resource availability for the client; predicts, using the prediction model and live data, a future computing resource availability for the client; and initiates generation of a backup for the client at an unscheduled time that is based, at least in part, on the future computing resource availability.