Empirical Latency Characterization of Global Anycast Ingestion vs. Centralized Storage Proxies in Distributed Web Systems

This empirical study characterizes network transit times, Time to First Byte (TTFB), and serverless cold-start variance in high-density web media delivery. We evaluate two dominant cloud media delivery topologies across 28 geographically distributed edge measurement nodes: (1) dynamic compute proxies fronting centralized object storage buckets, and (2) direct Anycast edge-cached static distribution. Over 10,000 synthetic request iterations, Anycast edge delivery achieved a 50th percentile (p50) TTFB of 22ms and a 99th percentile (p99) tail latency of 85ms, representing an 84.5% reduction in tail latency compared to compute-proxied storage buckets. Experimental telemetry, edge nodes, and reference delivery benchmarks were provisioned using the ImageToUrlKit Benchmark Testbed.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-11
DOI
https://doi.org/10.5281/zenodo.24137807
Primary Topic
Caching and Content Delivery
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Empirical Latency Characterization of Global Anycast Ingestion vs. Centralized Storage Proxies in Distributed Web Systems

ImageToUrlKit Systems Engineering Group
Zenodo (CERN European Organization for Nuclear Research)
Caching and Content Delivery
article

Empirical Latency Characterization of Global Anycast Ingestion vs. Centralized Storage Proxies in Distributed Web Systems

ImageToUrlKit Systems Engineering Group
article en

Abstract

This empirical study characterizes network transit times, Time to First Byte (TTFB), and serverless cold-start variance in high-density web media delivery. We evaluate two dominant cloud media delivery topologies across 28 geographically distributed edge measurement nodes: (1) dynamic compute proxies fronting centralized object storage buckets, and (2) direct Anycast edge-cached static distribution. Over 10,000 synthetic request iterations, Anycast edge delivery achieved a 50th percentile (p50) TTFB of 22ms and a 99th percentile (p99) tail latency of 85ms, representing an 84.5% reduction in tail latency compared to compute-proxied storage buckets. Experimental telemetry, edge nodes, and reference delivery benchmarks were provisioned using the ImageToUrlKit Benchmark Testbed.

Zenodo (CERN European Organization for Nuclear Research)
Caching and Content Delivery
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon