Global Site Selector for Seamless Application Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-datacenter deployments face challenges in seamless application hosting and migration due to asymmetric application distribution, leading to suboptimal load balancing and failover mechanisms that result in excessive overhead and potential multiple-hop request-response paths.
Innovation Solution
A system and method that utilize Global Site Selectors (GSS) and Application Control Engines (ACE) to forward HTTP requests to suitable datacenters, configuring URL match policies and VIP addresses to ensure seamless application hosting and migration, while avoiding URL redirects and excessive health monitoring, by synchronizing application configurations across datacenters and using enhanced keepalive protocols.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If asymmetric application distribution is used across multiple datacenters, then application hosting flexibility is improved, but load balancing efficiency deteriorates
Solution Approach 1:
The patent introduces a Global Site Selector (GSS) as an intermediary component that sits between clients and multiple datacenters. The GSS receives client requests, determines the optimal datacenter based on application availability and load conditions, and forwards requests accordingly. This mediator resolves the contradiction by enabling flexible asymmetric application distribution while maintaining efficient load balancing through centralized intelligence.
Solution Approach 2:
The system implements dynamic request routing where the GSS continuously monitors application distribution and load conditions across datacenters, adjusting routing decisions in real-time. This dynamic adaptation allows the system to handle asymmetric application distribution effectively while optimizing load balancing based on current conditions, resolving the contradiction between flexibility and efficiency.
2Reliability
If traditional failover mechanisms are implemented in multi-datacenter deployments, then system reliability is improved, but overhead increases
Solution Approach 1:
The patent implements a self-service health monitoring mechanism where each datacenter autonomously monitors its own application availability and reports status to the GSS. The GSS uses this information to make routing decisions without requiring complex centralized health checking infrastructure. This self-service approach improves reliability through continuous monitoring while reducing overhead by eliminating redundant health check protocols.
Solution Approach 2:
The system establishes a feedback loop where datacenters report their application availability status to the GSS, which then adjusts routing decisions accordingly. This feedback mechanism enables reliable failover by continuously monitoring system state and automatically rerouting requests when applications become unavailable, achieving high reliability without excessive overhead.
3Reliability
If multiple-hop request-response paths are used for application access, then fault tolerance is improved, but response time increases
Solution Approach 1:
The patent implements preliminary action by having the GSS pre-determine the optimal datacenter for each request based on application availability and client location before forwarding the request. This upfront decision-making eliminates the need for multiple-hop paths with retry logic, achieving fault tolerance through proactive routing while minimizing response time by avoiding unnecessary intermediate hops.
Solution Approach 2:
The system changes the routing parameter from static multi-hop paths to dynamic single-hop routing based on real-time application availability. The GSS modifies the request destination parameter to point directly to the optimal datacenter, maintaining fault tolerance through intelligent routing while reducing response time by eliminating intermediate hops.
Data Source
AI summary
An example method includes receiving a hypertext transfer protocol (HTTP) request from a client for an application at a first datacenter at which the application is not available; and querying a first global site selector (GSS) configured to identify a second datacenter at which the application is available. The second datacenter can be identified by a selected virtual Internet Protocol (VIP) address. The method can also include forwarding the HTTP request to the selected VIP address; and responding to the client with an HTTP response received from the selected VIP address.