Data Center Selection via Orthogonal Network Paths
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud services are unreliable due to network outages, as existing systems fail to effectively select diverse data centers for service delivery, leading to single points of failure and service unavailability.
Innovation Solution
A system that uses a network-aware data center selection coordinator to intelligently select a set of servicing data centers based on network path information, minimizing common network components and ensuring independence between paths, thereby providing primary and backup service instances for improved availability and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data centers are selected without considering network path diversity, then device complexity is reduced, but service reliability deteriorates due to single points of failure
Solution Approach 1:
The system performs preliminary analysis of network path information before selecting data centers. The selection coordinator obtains network path information from network elements and analyzes it in advance to identify data centers with independent network paths, preventing single points of failure before service deployment begins.
Solution Approach 2:
A selection coordinator acts as an intermediary between the user device and multiple data centers. This intermediary analyzes network path information and intelligently selects data centers with independent network paths, resolving the contradiction by adding a coordination layer that improves reliability without requiring complex client-side implementation.
2Reliability
If multiple data centers are selected to provide backup service instances, then service reliability is improved, but network path commonality increases leading to potential single points of failure
Solution Approach 1:
The system applies different selection criteria to different data centers based on their network path characteristics. Each data center is evaluated individually for network path independence, and the selection coordinator chooses data centers whose network paths diverge at different points, ensuring that local network failures do not affect multiple data centers simultaneously.
Solution Approach 2:
The system obtains and analyzes network path information from network elements before making data center selections. This preliminary analysis identifies data centers with independent network paths, allowing the system to select multiple data centers for redundancy while avoiding those that share common network components that could create single points of failure.
3Reliability
If network path analysis is performed for data center selection, then service availability is improved, but processing time and system complexity increase
Solution Approach 1:
The system obtains network path information from network elements in advance of service deployment. By performing this analysis preliminarily, the system has network path data ready when data center selection is needed, reducing the time required for selection while maintaining the reliability benefits of network path analysis.
Solution Approach 2:
Network elements provide their own network path information to the selection coordinator without requiring extensive external analysis. This self-service approach allows the system to obtain necessary network path data efficiently, reducing processing time while still enabling intelligent data center selection based on network path independence.
Data Source
AI summary
A capability is provided for improving service availability experienced by a user requesting access to a service via a user device, where the service has multiple service instances available from multiple data centers. A set of servicing data centers is selected for the user device. The set of servicing data centers is selected from a set of candidate data centers. The set of servicing data centers is selected from the set of candidate data centers in a manner for maximizing the independence/orthogonality of the network paths between the user device and the servicing data centers. The set of servicing data centers includes two or more of the candidate data centers from which the user device may access the service, thereby providing the user device with a primary service instance and one or more backup service instances which may be used to access the service.