Shared Network Path Assurance with Hash-Based Probing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional network monitoring in ECMP environments fails to ensure compliance with service level agreements (SLAs) due to averaging network metrics, which can conceal issues in specific segments, and requires numerous probes to assess individual path performance accurately.
Innovation Solution
Implementing a method to identify and monitor each distinct path in a network by using a collection of hashes, where each hash corresponds to a specific path, and transmitting probes with these hashes to determine network performance, allowing for more accurate measurement of individual path metrics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If conventional network monitoring averages network metrics across all paths, then the monitoring process is simple, but it conceals performance issues in specific path segments
Solution Approach 1:
The patent segments the network monitoring by dividing paths into distinct segments or hops. Each segment is monitored individually using segment routing identifiers, allowing precise identification of performance issues in specific path segments rather than averaging across the entire path. This segmentation enables granular measurement of latency, packet loss, and other metrics at each hop.
Solution Approach 2:
The patent applies local quality by assigning unique identifiers to specific path segments and monitoring each segment with tailored metrics. Instead of uniform monitoring across all paths, the system captures locally-specific performance characteristics of each segment, enabling precise identification of where performance degradation occurs in the network topology.
2Measurement precision
If numerous probes are transmitted to assess individual path performance accurately, then measurement precision improves, but the quantity of probes and system resources increase
Solution Approach 1:
The patent extracts path identification information from the packet header using segment routing identifiers. By embedding and reading these identifiers directly from packet headers, the system can track individual path performance without transmitting separate probes for each path assessment. This extraction approach reduces probe overhead while maintaining measurement precision.
Solution Approach 2:
The patent implements self-service monitoring where packets carry their own path identification information through segment routing identifiers embedded in their headers. Each packet essentially monitors its own path by including its identifier, which is then read by network nodes to track performance metrics without requiring external probing infrastructure.
3Productivity
If path selection is made using local hashing mechanism with multiple key values, then load balancing efficiency improves, but the complexity of path identification and monitoring increases
Solution Approach 1:
The patent applies preliminary action by pre-computing and embedding segment routing identifiers in packet headers before transmission. This preliminary encoding of path information allows transit nodes to efficiently perform load balancing using hashing mechanisms while simultaneously enabling straightforward path identification and monitoring without requiring complex real-time computation or additional monitoring infrastructure.
Data Source
AI summary
Disclosed are systems, apparatuses, methods, and computer-readable media to measure performance of distinct paths of a network. A method includes determining a collection of hashes of a network based on a network probe event, each hash in the collection of hashes corresponding to a distinct path from a first edge device to a second edge device through the network; transmitting a collection of probes from the first edge device in the network, wherein each probe in the collection of probes is assigned a hash selected from the collection of hashes; receiving probes from the collection of probes at the second edge device; and determining a network performance of each distinct path through the network.