Cloud Failover Resource Scaling for Idle Capacity Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Deploying primary and backup instances of mission critical applications in geographically dispersed cloud environments leads to unnecessary doubling of computing resources, increasing costs, power consumption, and cooling requirements due to idle resources.
Innovation Solution
Implement a primary application cloud instance that allocates necessary resources based on historical usage and provides heartbeat and session state information to a backup instance, which in turn allocates a minimum quantity of resources for a skeletal version, optimizing resource usage and enabling swift failover.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a backup instance is deployed in a geographically dispersed location for fault tolerance, then system reliability is improved, but computing resources are doubled and half of the assigned computing resources remain idle
Solution Approach 1:
The backup instance dynamically adjusts its resource allocation based on operational status. During normal operation, it maintains minimal resource usage, but automatically scales up to full capacity when failing over to become the primary instance, eliminating the need for permanently allocated idle resources
Solution Approach 2:
The system changes the operational parameters of the backup instance between two states: a dormant state with minimal resource consumption and an active state with full resource allocation. This parameter switching allows the system to maintain reliability while optimizing resource utilization
2Reliability
If idle computing resources are maintained for backup instances, then failover capability is ensured, but power consumption increases unnecessarily
Solution Approach 1:
The backup instance performs preliminary setup and configuration in advance, maintaining only essential core services in a dormant state. When failover is required, the remaining services are rapidly activated, achieving both reliability and energy efficiency by avoiding continuous full-power operation
Solution Approach 2:
The backup instance autonomously manages its own resource allocation and activation, transitioning from a low-power dormant state to a fully operational state only when needed, thereby eliminating unnecessary power consumption while maintaining failover capability
3Speed
If full computing resources are allocated to backup instances, then rapid failover is enabled, but cooling requirements and operational costs increase
Solution Approach 1:
The system dynamically adjusts resource allocation based on real-time needs, maintaining minimal resources during normal operation and rapidly scaling to full capacity during failover. This dynamic approach achieves rapid failover while minimizing continuous cooling requirements
Data Source
AI summary
A primary application cloud instance may receive historical usage of an application, may allocate, based on the historical usage, a quantity of cloud resources for enabling the application to be accessed, and may enable the application to be accessed. The primary application cloud instance may provide, to a backup application cloud instance that allocates a minimum quantity of cloud resources for providing a skeletal version of the application, heartbeat and session state information associated with the primary application cloud instance, and the quantity of cloud resources for enabling the application to be accessed. The primary application cloud instance may provide, to the backup application cloud instance, an indication of a failure of the primary application cloud instance, via the heartbeat and session state information, to cause the backup application cloud instance to allocate the quantity of cloud resources and to enable access to the application.