Selected Publications
- Anup Agarwal, Venkat Arun, Devdeep Ray, Ruben Martins, and Srinivasan Seshan
@inproceedings{ccmatic-nsdi, author = {Agarwal, Anup and Arun, Venkat and Ray, Devdeep and Martins, Ruben and Seshan, Srinivasan}, title = {Towards provably performant congestion control}, booktitle = {21st {USENIX} Symposium on Networked Systems Design and Implementation ({NSDI} 24)}, year = {2024}, address = {Santa Clara, CA}, conference = {USENIX NSDI}, month = apr, publisher = {{USENIX} Association} } - Anup Agarwal, Shadi Noghabi, Inigo Goiri, Srinivasan Seshan, and Anirudh Badam
@inproceedings{slacksched, author = {Agarwal, Anup and Noghabi, Shadi and Goiri, Inigo and Seshan, Srinivasan and Badam, Anirudh}, title = {Unlocking unallocated cloud capacity for long, uninterruptible workloads}, booktitle = {20th USENIX Symposium on Networked Systems Design and Implementation (NSDI 23)}, year = {2023}, isbn = {978-1-939133-33-5}, address = {Boston, MA}, pages = {457--478}, url = {https://www.usenix.org/conference/nsdi23/presentation/agarwal-anup}, publisher = {USENIX Association}, month = apr, conference = {USENIX NSDI} } - Anup Agarwal, Venkat Arun, Devdeep Ray, Ruben Martins, and Srinivasan Seshan
Heuristics are ubiquitous in computer systems. Examples include congestion control, adaptive bit rate streaming, scheduling, load balancing, and caching. In some domains, theoretical proofs have provided clarity on the conditions where a heuristic is guaranteed to work well. This has not been possible in all domains because proving such guarantees can involve combinatorial reasoning making it hard, cumbersome and error-prone. In this paper we argue that computers should help humans with the combinatorial part of reasoning. We model reasoning questions as āā formulas [1] and solve them using the counterexample guided inductive synthesis (CEGIS) framework. As preliminary evidence, we prototype CCmatic, a tool that semi-automatically synthesizes congestion control algorithms that are provably robust. It rediscovered a recent congestion control algorithm that provably achieves high utilization and bounded delay under a challenging network model. It also found previously unknown variants of the algorithm that achieve different throughput-delay trade-offs.
@inproceedings{ccmatic, author = {Agarwal, Anup and Arun, Venkat and Ray, Devdeep and Martins, Ruben and Seshan, Srinivasan}, title = {Automating network heuristic design and analysis}, year = {2022}, isbn = {9781450398992}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3563766.3564085}, doi = {10.1145/3563766.3564085}, booktitle = {Proceedings of the 21st ACM Workshop on Hot Topics in Networks}, pages = {8ā16}, numpages = {9}, keywords = {automated reasoning, congestion control}, location = {Austin, Texas}, series = {HotNets '22}, conference = {HotNets} } - Anup Agarwal, Zaoxing Liu, and Srinivasan Seshan
@inproceedings{heterosketch, author = {Agarwal, Anup and Liu, Zaoxing and Seshan, Srinivasan}, title = {HeteroSketch: Coordinating Network-wide Monitoring in Hetero-geneous and Dynamic Networks}, booktitle = {19th {USENIX} Symposium on Networked Systems Design and Implementation ({NSDI} 22)}, year = {2022}, url = {https://www.usenix.org/conference/nsdi22/presentation/agarwal}, address = {Renton, WA}, conference = {USENIX NSDI}, month = apr, publisher = {{USENIX} Association} } - Anup Agarwal, Jinghan Sun, Shadi Noghabi, Srinivasan Iyengar, Anirudh Badam, Ranveer Chandra, Srinivasan Seshan, and Shivkumar Kalyanaraman
Renewable energy is becoming an important power source for data centers, especially with the zero-carbon waste pledges made by big cloud providers. However, one of the main challenges of renewable energy sources is the high variability of power produced. Traditional approaches such as batteries or transmitting to the grid fall short on scale, overhead, or "green-ness". We propose Virtual Battery: instead of adapting the availability of power to match the computation demand we shift computational demand to meet the availability of power. Virtual batteries shift demand by requiring applications to either be flexible and delay-tolerant or proactively migrating to where power is (going to be) available. We show that using multiple virtual battery sites in combination can meet the needs of modern applications. Moreover, we show how an intelligent network and power aware co-scheduler can not only provide availability despite variability but also help mitigate migration related network overhead by over 30 % in total and 4.2x at peak.
@inproceedings{vb, author = {Agarwal, Anup and Sun, Jinghan and Noghabi, Shadi and Iyengar, Srinivasan and Badam, Anirudh and Chandra, Ranveer and Seshan, Srinivasan and Kalyanaraman, Shivkumar}, title = {Redesigning Data Centers for Renewable Energy}, booktitle = {Proceedings of the Twentieth ACM Workshop on Hot Topics in Networks}, year = {2021}, pages = {45-52}, doi = {10.1145/3484266.3487394}, url = {https://doi.org/10.1145/3484266.3487394}, address = {New York, NY, USA}, isbn = {9781450390873}, location = {Virtual Event, United Kingdom}, numpages = {8}, publisher = {Association for Computing Machinery}, series = {HotNets '21}, conference = {HotNets} } - Prateesh Goyal, Anup Agarwal, Ravi Netravali, Mohammad Alizadeh, and Hari Balakrishnan
@inproceedings{abc, author = {Goyal, Prateesh and Agarwal, Anup and Netravali, Ravi and Alizadeh, Mohammad and Balakrishnan, Hari}, title = {{ABC}: A Simple Explicit Congestion Controller for Wireless Networks }, booktitle = {17th {USENIX} Symposium on Networked Systems Design and Implementation ({NSDI} 20)}, year = {2020}, pages = {353--372}, url = {https://www.usenix.org/conference/nsdi20/presentation/goyal}, address = {Santa Clara, CA}, conference = {USENIX NSDI}, isbn = {978-1-939133-13-7}, month = feb, publisher = {{USENIX} Association} }