Distributed Disaster Recovery for Virtual Computing Services
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current disaster recovery systems for virtual computing services are inefficient due to high overhead costs and insufficient capacity, as they require dedicated backup data centers that are often idle and can become single points of failure, making it difficult to transition multiple customers to a backup site during a disaster.
Innovation Solution
A distributed disaster recovery system where each enterprise customer's transaction data and configuration data are mirrored across multiple physical locations, allowing any secondary location to provide virtual computing services in case of a disaster, with each location having less than 100% reserve capacity and sufficient resources to meet contracted service levels, and using a configuration replication engine to manage data replication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a dedicated backup data center is established for each enterprise customer, then disaster recovery reliability is improved, but resource overhead and cost increase to 100% for many resources
Solution Approach 1:
The patent combines disaster recovery resources across multiple data centers to serve multiple customers collectively. Instead of each customer having dedicated backup resources, the system merges backup capabilities into shared pools at different data centers, allowing resources to be utilized by multiple customers and reducing overall resource overhead while maintaining disaster recovery reliability
Solution Approach 2:
The patent makes data center resources universal by enabling each data center to serve both as a primary production site for some customers and as a disaster recovery site for other customers. This multi-functionality allows resources to be dynamically allocated based on need, eliminating the waste of dedicated backup resources that remain idle during normal operations
2Quantity of substance
If a central shared backup data center is established to serve multiple customers, then resource efficiency is improved, but the backup site becomes a single point of failure and capacity may be insufficient during major disasters
Solution Approach 1:
The patent segments the disaster recovery function across multiple data centers rather than consolidating it into a single central backup facility. Each data center maintains some reserve capacity for disaster recovery, and customers are assigned to different backup locations, distributing the risk and ensuring that a single point of failure does not affect all customers
Solution Approach 2:
The patent applies local quality by giving each data center specific disaster recovery capabilities tailored to the needs of its assigned customers. Rather than a uniform central backup, each location has localized backup resources and capacity appropriate to the customers it serves, ensuring sufficient capacity at each site while maintaining overall resource efficiency
3Reliability
If a dedicated backup data center is established, then disaster recovery capability is improved, but transition complexity increases when multiple customers need to move to backup simultaneously
Solution Approach 1:
The patent implements preliminary action by pre-configuring backup resources and establishing backup relationships between data centers before disasters occur. Configuration data and transaction data are replicated in advance, and backup capacity is reserved at each location, enabling rapid transition without complex real-time coordination when multiple customers need to move simultaneously
Data Source
AI summary
Systems and methods for providing disaster recovery services to enterprise customers of virtual computing services are described. Disaster recovery services may be provided in data centers that also provide primary computing services to enterprise customers. Disaster recovery services may be distributed across the data centers, such that no dedicated disaster recovery site exists. In addition, disaster recovery services may be provided such that each of the enterprise customers receiving primary services at a given data center may be distributed across the remaining data centers for disaster recovery purposes.