Named Capacity Reservation Pools for Cloud Resource Allocation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In large-scale data centers, managing network-accessible resources becomes complex due to high utilization levels and resource failures, making it difficult for clients to acquire necessary resources during spikes in workload or failures, especially when long-term reservations are affected.

Innovation Solution

Implementing a resource manager that allows clients to create named pools of reserved instances dedicated to specific purposes, enabling flexible capacity management, billing, and account linkage for efficient resource allocation and utilization.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If long-term resource reservations are made to ensure availability, then reliability is improved, but flexibility and adaptability deteriorate when workload spikes or failures occur

Engineering Contradiction:
Improveresource availabilityVSAvoidcapacity flexibility
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The patent segments reserved instances into named pools organized by application or service type. Each pool is independent and can be managed separately, allowing customers to allocate resources to specific applications without affecting other applications. This segmentation enables flexible capacity management while maintaining reservation guarantees, resolving the contradiction between reliability and adaptability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces dynamic pool management capabilities where customers can modify pool specifications, transfer instances between pools, and adjust capacity allocations in response to changing workloads. The system dynamically adapts to workload variations while maintaining the underlying reservation guarantees, enabling both reliability and flexibility.

Inventive Principle:
Principle #15Dynamics

2Reliability

If reserved instances are allocated to ensure dedicated capacity, then reliability is improved, but device complexity increases due to multiple resource acquisition modes

Engineering Contradiction:
Improvededicated capacityVSAvoidresource management complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent creates a unified resource management framework that handles multiple resource acquisition modes (reserved instances, on-demand instances, spot instances) through a single named pools interface. This universal approach simplifies resource management by providing consistent pool-based organization and management across different acquisition methods, reducing complexity while maintaining dedicated capacity guarantees.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Adaptability or versatility

If multiple resource acquisition modes are supported to meet diverse needs, then adaptability is improved, but ease of operation deteriorates due to complex provisioning and administration

Engineering Contradiction:
Improveresource acquisition flexibilityVSAvoidprovisioning simplicity
Core Design Contradiction:
Adaptability or versatilityVSEase of operation

Solution Approach 1:

By segmenting resources into named pools organized by application or service type, the patent simplifies provisioning and administration. Customers can create pools with specific capacity requirements and manage them independently, making the system easier to operate while maintaining the ability to acquire resources through multiple modes.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The named pools structure acts as an intermediary layer between customers and the underlying resource acquisition mechanisms. This intermediary simplifies the interface for customers while abstracting the complexity of multiple acquisition modes, making the system both adaptable and easy to operate.

Inventive Principle:
Principle #24Intermediary (Mediator)

4Productivity

If capacity is reserved for expected workload levels, then productivity is improved, but loss of time increases when actual workload exceeds expectations during high-utilization periods

Engineering Contradiction:
Improveworkload handling capacityVSAvoidresource acquisition delay
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent enables customers to reserve capacity in advance through named pools for expected workload levels. When workloads spike, customers can quickly access additional capacity from their reserved pools or transition to on-demand/spot instances without lengthy provisioning delays, reducing time loss while maintaining productivity.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The dynamic pool management system allows customers to adapt their capacity allocation in real-time based on actual workload conditions. When workloads exceed expectations, the system can dynamically adjust pool capacities or switch between acquisition modes, minimizing downtime and maintaining productivity.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS10152449B1User-defined capacity reservation pools for network-accessible resources
Publication Date: 2018.12.11 AMAZON TECH INC
  • US10152449B1 patent drawing
  • US10152449B1 patent drawing
  • US10152449B1 patent drawing

AI summary

Methods and apparatus for managing user-defined capacity reservation pools for network-accessible resources are disclosed. A system includes a plurality of resource instances of a provider network, and a resource manager. In response to a pool assignment request from a client specifying a pool identifier of a reserved instance pool created on behalf of the client, the resource manager assigns one or more resource instance reservations of the client to a pool with the first pool identifier. In response to an instance activation request from the client, the resource manager identifies a particular pool created on behalf of the client from which to select a particular resource instance reservation for activation, and activates the particular resource instance.