Online services all seek to provide their customers with the best Quality of Experience (QoE) possible. Milliseconds of delay can cause users to abandon a cat video or move onto a different shopping site, which translates into lost revenue. Thus, minimizing latency between users and content is crucial. To reduce latency, content and cloud providers have built massive, global networks. However, their networks must interact with customer ISPs via BGP, which has no concept of performance. The shortcomings of BGP are many and well documented, but in this paper we ask the community to take a step back and rethink what we know about BGP. We examine three separate studies of performance using large content and cloud provider networks and find that performance-aware routing schemes rarely achieve lower latency than BGP. We lay out a map for research to further study the idea that beating BGP may be more difficult than previously thought.
|Original language||American English|
|Title of host publication||HotNets 2019 - Proceedings of the 18th ACM Workshop on Hot Topics in Networks|
|Publisher||Association for Computing Machinery, Inc|
|Number of pages||8|
|State||Published - 13 Nov 2019|
|Event||18th ACM Workshop on Hot Topics in Networks, HotNets 2019 - Princeton, United States|
Duration: 14 Nov 2019 → 15 Nov 2019
|Name||HotNets 2019 - Proceedings of the 18th ACM Workshop on Hot Topics in Networks|
|Conference||18th ACM Workshop on Hot Topics in Networks, HotNets 2019|
|Period||14/11/19 → 15/11/19|
Bibliographical noteFunding Information:
Acknowledgements. We appreciate the valuable feedback from the HotNets reviewers. We acknowledge the contributions of our collaborators on studies that informed this paper. This work was partly funded by NSF awards CNS-1835253, CNS-1835252, CNS-1413978, and CNS-1563849.
© 2019 ACM.
- Content delivery
- Traffic engineering