SaaS Probe Aggregation in Router Clusters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for probing Software-as-a-Service (SaaS) clouds result in redundant probes across multiple network devices using the same internet service provider, leading to excessive network traffic and inefficiencies, especially in scaled solutions with multiple hub routers.
Innovation Solution
Distributing probing responsibilities among multiple routers in a colocation facility and sharing probe results across the cluster, allowing each router to determine the best path without individually probing every path, thereby reducing redundant traffic and optimizing performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If probes are enabled for the same set of applications on each network device, then routing performance data can be obtained for path selection, but redundant probes generate excessive network traffic
Solution Approach 1:
The patent merges probing operations across multiple network devices by having them share a common probe result database. Instead of each device independently probing applications, devices collaborate to generate and share probe results, eliminating redundant traffic while maintaining comprehensive routing performance data across the network.
Solution Approach 2:
The patent implements a universal probe result database that serves all network devices in the group. This shared database allows any device to benefit from probe results generated by any other device, making the probing system multi-functional and eliminating the need for duplicate probing operations on each device.
2Reliability
If all network devices perform individual probing, then each device has complete routing information, but the system complexity and overhead increase significantly
Solution Approach 1:
The patent segments the probing function by designating specific devices as probe initiators for specific applications or time periods, while other devices act as consumers of probe results. This segmentation reduces individual device complexity and overall system overhead while maintaining complete routing information availability through the shared database.
Solution Approach 2:
The patent implements a self-service mechanism where the probe result database automatically stores, manages, and distributes routing performance data to all devices. This eliminates the need for each device to independently manage its own probing operations and data storage, significantly reducing system overhead while maintaining information completeness.
3Measurement precision
If probes are sent frequently to ensure accurate performance data, then routing decisions are optimized, but network traffic and processing load increase
Solution Approach 1:
The patent implements periodic probing where devices send probes at optimized intervals rather than continuously. The probe result database stores historical data and allows devices to make routing decisions based on both recent and historical performance data, maintaining accuracy while reducing the frequency of probe transmissions and improving network efficiency.
Data Source
AI summary
Techniques for sharing the probing of software-as-a-service clouds among a cluster of routers are described herein. The techniques may include establishing a first path between a cluster of routers and an application infrastructure. Establishing a second path between the cluster of routers and the application infrastructure. Designating a first router in the cluster of routers to send probes over the first path to the application infrastructure. Designating a second router in the cluster of routers to send probes over the second path to the application infrastructure. Distributing, by the first router and to the cluster of routers, first routing performance data indicating a performance of the first path when communicating with the application infrastructure over the first path, distributing, by the second router and to the cluster of routers, second routing performance data indicating a performance of the second path when communicating with the application infrastructure over the second path.