Compute Unit Spread Scoring for Failure Correlation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In cloud computing systems, existing technologies lack effective methods to manage and optimize the placement of computing units to minimize the risk of correlated failures across different units, leading to potential service disruptions and loss of flexibility in resource management.

Innovation Solution

The provisioning application in the cloud computing system uses a spread score to assess the resilience of computing units and allocates them based on failure correlation data, ensuring geographic and network topology diversity to reduce the impact of failures, while allowing customers to control and optimize their resource distribution.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If computing units are placed close together to improve resource utilization, then resource efficiency increases, but the risk of correlated failures increases

Engineering Contradiction:

Improveresource utilizationVSAvoidfailure correlation risk

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system segments computing units into different groups based on their failure correlation characteristics. By dividing the infrastructure into independent segments (different data centers, availability zones, or physical locations), the system can place computing units close together for resource efficiency while ensuring that failures in one segment do not propagate to other segments, thus resolving the contradiction between resource utilization and failure correlation risk.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system applies different placement strategies to different computing units based on their specific requirements and the local infrastructure characteristics. By evaluating failure correlation data for each computing unit and its potential placement location, the system can optimize resource utilization locally while maintaining overall system reliability through diversified placement across multiple locations with different failure profiles.

Inventive Principle:

Principle #3Local quality

2Reliability

If computing units are distributed across multiple locations to reduce failure correlation, then system resilience improves, but resource management complexity increases

Engineering Contradiction:

Improvesystem resilienceVSAvoidresource management complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system implements automated resource management that self-adjusts computing unit placements based on real-time failure correlation analysis. The resource manager automatically evaluates placement options, selects optimal locations that minimize failure correlation, and reallocates computing units as needed without requiring manual intervention. This automation resolves the contradiction by maintaining high system resilience through distributed placement while eliminating the burden of manual resource management complexity.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system continuously monitors infrastructure health, failure patterns, and computing unit performance, using this feedback to dynamically adjust placement decisions. By incorporating real-time feedback loops that analyze failure correlation data and automatically reposition computing units when risk thresholds are exceeded, the system maintains high resilience while managing complexity through data-driven automation rather than static, manually-configured distributions.

Inventive Principle:

Principle #23Feedback

3Adaptability or versatility

If physical location information is concealed from customers to maintain flexibility, then operator flexibility improves, but customer transparency deteriorates

Engineering Contradiction:

Improveoperator flexibilityVSAvoidcustomer transparency

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The system introduces an intermediary layer (virtualization layer or abstraction layer) between the customer and the physical infrastructure. This intermediary allows the operator to conceal specific physical location details from customers while maintaining full flexibility in resource management. The intermediary presents a standardized, location-agnostic interface to customers while the backend can freely relocate computing units based on failure correlation analysis, thus resolving the contradiction by decoupling customer visibility from physical placement flexibility.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS8966037B1Measuring spread of compute capacity

Publication Date: 2015.02.24 AMAZON TECH INC

AI summary

Disclosed are various embodiments for provisioning computing units. A spread request is received. The spread request relates to a class of assigned computing units residing within a plurality of networked computing units. The spread request is associated with a spread criteria. In response to the request, a plurality of networked computing units is provisioned based on failure correlation data and in accordance with the spread criteria, to produce a final spread score. Success is indicated in response to the request if the final spread score meets the spread criteria.