Sequoia: Treeness of Internet Path Metrics
Sequoia aims to make distributed applications network-aware. That is, enable applications to take advantage of the characteristics of the underlying network such as proximity, bandwidth capacity, and topology. It intends to achieve this through the key concept of prediction trees, a virtual topology of the network, where virtual nodes representing routers connect real end hosts, and carefully computed edge weights model path properties such as latency and bandwidth.
The following are some of the network-aware functionalities that Sequoia seeks to provide:
- Path Property Prediction: Two hosts in the system can estimate path properties (latency, loss rate, available bandwidth) between them through computation without requiring a direct measurement.
- Server Selection: An external client outside the system can select the closest or best-provisioned server in the system.
- Hierarchical Partitioning: Hosts can organize into hierarchical clusters based on their position in the network.
Publications
- On the Treeness of Internet Latency and Bandwidth
Venugopalan Ramasubramanian, Dahlia Malkhi, Fabian Kuhn, Mahesh Balakrishnan, Archit Gupta and Aditya Akella.
ACM SIGMETRICS / Performance, Seattle WA, June 2009. - A Unified Network Coordinate System for Bandwidth and Latency
Venugopalan Ramasubramanian, Dahlia Malkhi, Fabian Kuhn, Ittai Abraham, Mahesh Balakrishnan, Archit Gupta and Aditya Akella.
Microsoft Research Technical Report MSR-TR-2008-124, September 2008. - The Internet Grows on Trees! (poster)
Dahlia Malkhi and Venugopalan Ramasubramanian.
Microsoft TechFest, Redmond, WA, March 2008. - Reconstructing Approximate Tree Metrics
Ittai Abraham, Mahesh Balakrishnan, Fabian Kuhn, Dahlia Malkhi, Venugopalan Ramasubramanian and Kunal Talwar.
ACM Symposium on Principles of Distributed Computing (PODC), Portland OR, August 2007. - Internet Path Quality Prediction Using Trees (poster)
Ittai Abraham, Mahesh Balakrishnan, Archit Gupta, Fabian Kuhn, Dahlia Malkhi, Venugopalan Ramasubramanian and Kunal Talwar.
Summer Intern Research Series, Microsoft Research Silicon Valley, July 2007.
Demo
A demo of geographical partitioning and closest-node discovery in Sequoia for PlanetLab nodes.
Talks
- Sequoia: Virtual-Tree Models for Internet Path Metrics (Rama)
Microsoft P2P Summit, Redmond, WA, March 2008.
ACM Symposium on Operating Systems Principles (SOSP), Stevenson WA, October 2007. (WIPS)
MSR Networking Mindswap, Kyoto, Japan, August 2007. - Reconstructing Approximate Tree Metrics (Fabian)
ACM Symposium on Principles of Distributed Computing (PODC), Portland OR, August 2007. - Sequoia: Supporting Latency-Aware Applications through Prediction Trees (Dahlia)
Research Workshop on Flexible Network Design, Bertinoro, Italy, October 2006.
People
- Ittai Abraham
- Mahesh Balakrishnan
- Dahlia Malkhi
- Venugopalan "Rama" Ramasubramanian
- Kunal Talwar
- Aditya Akella (visiting researcher)
- Fabian Kuhn (former post doc)
