DNS Reachability Monitoring for Multi-Path Application Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Assessing the network impact on application performance and user experience is challenging in cloud-hosted, geographically distributed environments, as existing network path monitoring is typically directed to the closest server/instance, leading to loss of connectivity and lack of visibility into secondary server performance during failures.
Innovation Solution
Implementing multi-path monitoring using DNS-based reachability information to discover and monitor multiple application instances, enabling proactive failover and traffic routing to secondary instances for enhanced user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If network path monitoring is directed to the closest server/instance via DNS resolution, then monitoring simplicity is maintained, but connectivity is lost and visibility into secondary servers is unavailable during failures
Solution Approach 1:
The system performs preliminary actions by discovering and storing multiple application instances (primary and secondary) before failure occurs. The monitoring agent proactively identifies alternate instances through DNS queries and prepares monitoring paths in advance, so when the primary instance fails, the system can immediately switch to a secondary instance without connectivity loss.
Solution Approach 2:
The DNS system serves as an intermediary that provides reachability information for multiple application instances. The monitoring agent uses DNS queries to discover primary and secondary instances, and the DNS responses act as a mediator that enables the system to switch between instances without direct knowledge of their identities or locations.
2Loss of information
If monitoring is directed only to the primary application instance, then monitoring focus is maintained, but visibility into secondary instance performance is lost
Solution Approach 1:
The system performs partial monitoring by focusing on the primary instance for detailed performance analysis while simultaneously performing excessive action by also monitoring secondary instances. This allows the system to maintain detailed visibility into primary instance performance while also gathering sufficient information about secondary instances to enable fast failover, balancing monitoring depth with failover capability.
Data Source
AI summary
In one embodiment, a device receives a domain name system text record containing Internet reachability information for a plurality of application instances of a particular application identified by a given domain name. The device determines a primary Internet address for a primary application instance of the plurality of application instances and one or more secondary Internet addresses for one or more respective secondary application instances of the plurality of application instances based on the Internet reachability information. The device configures a path monitoring agent to perform path performance monitoring of a network path to the primary Internet address and perform path performance monitoring of one or more alternate network paths to the one or more secondary Internet addresses.