How Cloud Edge Infrastructure Improves Network Performance in Europe

Data centers have become a central part of cloud service infrastructure.

Cloud providers typically build data centers in large metropolitan areas where most users live and work. These centralized data centers (cloud regions) serve as a regional backbone for cloud providers.

In some countries, cloud providers establish smaller, more distributed data centers. While these cloud-edge data have been shown to reduce latency by up to 55% for their local customers, because of their size, they provide a limited range of services.

Recently, Georgios Tsiknakis, Antonis Chatzivasiliou, and I at the University of Crete sought to validate previous continental cloud-edge performance studies conducted by Martin and Dogar, narrowing the focus to a country level.

We used Virtual Machines (VMs) established in each cloud region of Amazon, Google, and Azure (Figure 1) in Europe to measure the Round Trip Time (RTT) — a measure of network speed/latency — from RIPE Atlas probes for Bulgaria, Czechia, Finland, Germany, Greece, and Poland. We categorized the countries into two groups according to the availability of cloud regions, geographic diversity, and strategic significance. (Learn more about our methodology).

Countries with Cloud-Edge Infrastructure Have Better RTT

Unsurprisingly, countries with cloud regions consistently achieve RTTs of under 20 ms for 80% of probe tests (Figure 2). The good RTTs in the baseline for Germany and Poland are not solely due to cloud regions located in those countries but also because they are close to cloud regions in neighboring countries. A special case is Finland, which achieves low RTTs despite not having VMs close to it except its own. This indicates that Finland has excellent Internet connectivity to all of Europe.

In countries without VMs (Greece, Czechia, Bulgaria), cloud-edge infrastructure demonstrates a 50-70% performance advantage over the closest cloud region located in another country (Figure 3). Consequently, countries closer to cloud regions located in neighboring countries benefit more than countries further away. For example, in Figure 1, Czechia is closer to cloud regions in Germany and Poland than Bulgaria’s nearest cloud region (Poland). And Bulgaria is closer than Greece’s closest cloud region (Northern Italy) and thus achieves lower RTTs.

Expanding Study to Other Countries 

We recognize that our work has several limitations, including: 

• The measurements were conducted in only six countries over four days in May 2023, which may not capture broader temporal trends.
• Geolocating vantage points relied on MaxMind, but the lack of historical data required us to use geolocation information from 2024, potentially introducing inaccuracies.
• Additionally, a limited number of VMs constrained our ability to perform comprehensive baseline measurements. These factors may affect the precision of our results.

In the future, we will measure the performance in more European countries and take measurements from more cloud regions, ensuring a complete view of the impact of cloud edge across Europe. 

We are also considering investigating at a city scale, which could be useful for Smart city initiatives.

Contributors: Georgios Tsiknakis, Antonis Chatzivasiliou.

Adapted from the original post that first appeared on RIPE Labs Blog.

Katerina Lionta is a postgraduate student in the computer science department at the University of Crete and a visitor at the Foundation for Research and Technology – Hellas (FORTH). The General Secretariat of Telecommunications and Posts funded this work.

The views expressed by the authors of this blog are their own and do not necessarily reflect the views of the Internet Society.