Table of Contents
Fetching ...

GLIDS: A Global Latency Information Dissemination System

Cyrill Krähenbühl, Seyedali Tabaeiaghdaei, Simon Scherrer, Matthias Frei, Adrian Perrig

TL;DR

The paper tackles the challenge of latency-aware path selection in multipath, inter-domain networks by disseminating propagation latency rather than attempting to predict full end-to-end latency. It introduces GLIDS, a two-subsystem design that measures intra-domain propagation latency and inter-domain links and distributes this information through SCION's control plane with hierarchical dissemination. The work provides implementation and evaluation on SCIONLab, demonstrates improvements in delay-based congestion control fairness, and shows potential latency reductions for first-packet latency and large-scale DNS routing scenarios. The findings suggest propagation latency transparency can enable more efficient path probing, better congestion control, and practical deployment strategies, while addressing privacy and veracity considerations for real-world adoption.

Abstract

A recent advance in networking is the deployment of path-aware multipath network architectures, where network endpoints are given multiple network paths to send their data on. In this work, we tackle the challenge of selecting paths for latency-sensitive applications. Even today's path-aware networks, which are much smaller than the current Internet, already offer dozens and in several cases over a hundred paths to a given destination, making it impractical to measure all path latencies to find the lowest latency path. Furthermore, for short flows, performing latency measurements may not provide benefits as the flow may finish before completing the measurements. To overcome these issues, we argue that endpoints should be provided with a latency estimate before sending any packets, enabling latency-aware path choice for the first packet sent. As we cannot predict the end-to-end latency due to dynamically changing queuing delays, we measure and disseminate the propagation latency, enabling novel use cases and solving concrete problems in current network protocols. We present the Global Latency Information Dissemination System (GLIDS), which is a step toward global latency transparency through the dissemination of propagation latency information.

GLIDS: A Global Latency Information Dissemination System

TL;DR

The paper tackles the challenge of latency-aware path selection in multipath, inter-domain networks by disseminating propagation latency rather than attempting to predict full end-to-end latency. It introduces GLIDS, a two-subsystem design that measures intra-domain propagation latency and inter-domain links and distributes this information through SCION's control plane with hierarchical dissemination. The work provides implementation and evaluation on SCIONLab, demonstrates improvements in delay-based congestion control fairness, and shows potential latency reductions for first-packet latency and large-scale DNS routing scenarios. The findings suggest propagation latency transparency can enable more efficient path probing, better congestion control, and practical deployment strategies, while addressing privacy and veracity considerations for real-world adoption.

Abstract

A recent advance in networking is the deployment of path-aware multipath network architectures, where network endpoints are given multiple network paths to send their data on. In this work, we tackle the challenge of selecting paths for latency-sensitive applications. Even today's path-aware networks, which are much smaller than the current Internet, already offer dozens and in several cases over a hundred paths to a given destination, making it impractical to measure all path latencies to find the lowest latency path. Furthermore, for short flows, performing latency measurements may not provide benefits as the flow may finish before completing the measurements. To overcome these issues, we argue that endpoints should be provided with a latency estimate before sending any packets, enabling latency-aware path choice for the first packet sent. As we cannot predict the end-to-end latency due to dynamically changing queuing delays, we measure and disseminate the propagation latency, enabling novel use cases and solving concrete problems in current network protocols. We present the Global Latency Information Dissemination System (GLIDS), which is a step toward global latency transparency through the dissemination of propagation latency information.
Paper Structure (42 sections, 7 figures)

This paper contains 42 sections, 7 figures.

Figures (7)

  • Figure 1: A SCION topology consisting of three ISDs and several possible end-to-end paths.
  • Figure 2: The number of distinct paths to all reachable ASes in the commercial SCION network from a single vantage point.
  • Figure 3: Topology where four hosts ($\text{H}_{1-4}$) combine path segments to end-to-end paths and leverage GLIDS to select paths. Diagonal lines indicate core ASes and core path segments.
  • Figure 4: Difference between the estimated and the experienced propagation latency in SCIONLab.
  • Figure 5: Latency reduction (in relative values or ms) when selecting the path with the lowest propagation latency.
  • ...and 2 more figures