An empirical study by Interconnection giant, DE-CIX, comparing and quantifying the performance differences between transit and peering interconnection strategies.
Whether you are a small hosting provider, an ISP, a content publisher, or a content application provider, at one point or another in your network’s growth cycle, you will ask yourself the question: When should I peer and when should I use transit? As soon as you have an ASN and are multihomed, you will think about peering.
The benefits of peering have been stated countless times. Peering reduces latency, improves the network quality, enhances the end user experience, reduces networks costs, etc. But what does that mean exactly?
An empirical study conducted by Adnan Ahmed, Zubair Shafiq, Harkeerat Bedi, and Amir Khakpour involved a large-scale measurement-based performance comparison of peering and transit interconnection strategies, quantifying the performance differences between transit and peering.
Peering vs. transit – A performance comparison of peering and transit interconnections
Example: CDNs and access ISPs
While content publishers and Content Delivery Networks traditionally relied on buying transit to deliver their content to end users, today they are deciding more and more in favor of peering relationships with access ISPs (or eyeball networks) to reduce transit costs.
Global Internet traffic volume has increased by more than 40% every year in recent years, and access ISPs are engaging increasingly in peering relationships instead of buying transit.
Peering paths substantially outperform transit paths
The results of the above-mentioned study show that:
- Peering paths outperform transit paths for 91% of Autonomous Systems (ASes);
- Peering paths have smaller propagation delays as compared to transit paths for more than 95% of ASes;
- Peering paths outperform transit paths in terms of propagation delay due to shorter path lengths;
Read the full paper at DE-CIX.net
DOWNLOAD WHITE PAPER