[Podcast] Is my Internet Down?
Research area: Network Analysis and Reasoning
Published by Ege Cem Kırcı on February 2, 2023
Originally written for the APNIC Blog
In this episode of PING, Ege Cem Kirci from the Networked Systems Group at ETH Zurich discusses his IMC paper on user perceptions of Internet outages, measured with Google Trends. This was presented at the IMC Conference 2022 in Nice, France in October 2022.
Ege and his co-authors have been exploring the relationship of Internet outages (for example, caused by weather events) and the information in Google Trends, with a mechanism to combine snapshots of data by time period which auto-scale, to a single unified time series. In their IMC talk, they presented SIFT (github), a detection and analysis tool for capturing user-affecting Internet outages. SIFT leverages user aggregated web search activity to detect outages.
The abstract from the IMC conference paper is a great overview of Ege’s work:
What are the worst outages for Internet users? How long do they last, and how wide are they? Such questions are hard to answer via traditional outage detection and analysis techniques, as they conventionally rely on network-level signals and do not necessarily represent users’ perceptions of connectivity.
We present SIFT, a detection and analysis tool for capturing user-affecting Internet outages. SIFT leverages users’ aggregated web search activity to detect outages. Specifically, SIFT starts by building a timeline of users’ interests in outage-related search queries. It then analyses this timeline looking for spikes of user interest. Finally, SIFT characterizes these spikes in duration, geographical extent, and simultaneously trending search terms that may help understand root causes, such as power outages or associated ISPs.
We use SIFT to collect more than 49,000 Internet outages in the United States over the last two years. Among others, SIFT reveals that user-affecting outages:
- Do not happen uniformly — half of them originate from 10 states only.
- Can affect users for a long time — 10% of them last at least 3 hours.
- Can have a broad impact — 11% of them simultaneously affect at least 10 distinct states.
SIFT annotations also reveal a perhaps overlooked fact — outages are often caused by climate and/or power-related issues.