Starvation ratio: letting applications drive datacenter congestion control

• Anderson Henrique da Silva Marcondes UDESC
• Enzo Bello Boscatto UDESC
• Guilherme Piêgas Koslovski UDESC

DOI: https://doi.org/10.5753/semish.2026.23208

Resumo

Datacenter performance is often limited by network-centric congestion controls relying on low-level metrics (e.g., packet loss, latency) that misinterpret applications needs. This work argues that applications should participate in congestion control decisions and introduces the starvation ratio (SR), a metric that detects when applications are truly limited by the network. Experimental evaluations within an 11-flow bottleneck scenario on a Linux-based prototype show that asynchronous applications can absorb network variations within a newly identified “silence zone” without degradation, proving conventional controls are overly restrictive. By deploying proactive and reactive mechanisms, our approach consistently reduces Flow Completion Time (FCT) for network-sensitive workloads. Notably, the proactive configuration eliminates micro-recovery delays, keeping the starvation ratio close to zero and reinforcing the baseline protocol through stable congestion window regulation. We conclude that shifting to application-driven signaling aligns network transmission with the receiver’s processing pace, preventing computational underutilization.

Referências

Agarwal, S., Krishnamurthy, A., and Agarwal, R. (2023). Host congestion control. In Proc. of the ACM SIGCOMM 2023 Conference, page 275–287. ACM.

Alizadeh, M., Greenberg, A., Maltz, D. A., Padhye, J., Patel, P., Prabhakar, B., Sengupta, S., and Sridharan, M. (2010). Data center tcp (dctcp). In Proc. of the ACM SIGCOMM 2010 Conference, pages 63–74.

Benson, T., Akella, A., and Maltz, D. A. (2010). Network traffic characteristics of data centers in the wild. In Proc. of the 10th ACM SIGCOMM Conference on Internet Measurement, IMC ’10, page 267–280. ACM.

Bernárdez, G., Suárez-Varela, J., Shi, X., Xiao, S., Cheng, X., Barlet-Ros, P., and Cabellos-Aparicio, A. (2025). GraphCC: a practical graph learning-based approach to congestion control in datacenters. Computer Networks, 257:110981.

Bonato, T., Kabbani, A., De Sensi, D., Pan, R., Le, Y., Raiciu, C., Handley, M., Schneider, T., Blach, N., Ghalayini, A., et al. (2024). FASTFLOW: flexible adaptive congestion control for high-performance datacenters. arXiv preprint arXiv:2404.01630.

Diel, G., Miers, C. C., Pillon, M. A., and Koslovski, G. P. (2023). RSCAT: Towards zero touch congestion control based on actor–critic reinforcement learning and software-defined networking. Journal of Network and Computer Applications, 215:103639.

Dukkipati, N. and McKeown, N. (2006). Why flow-completion time is the right metric for congestion control. ACM SIGCOMM Computer Communication Review, 36(1):59–62.

Ewais, M. and Chow, P. (2023). Disaggregated memory in the datacenter: a survey. IEEE Access, 11:20688–20712.

Geng, Y., Zhang, H., Shi, X., Wang, J., Yin, X., He, D., and Li, Y. (2023). Delay based congestion control for cross-datacenter networks. In 2023 IEEE/ACM 31st International Symposium on Quality of Service (IWQoS), pages 1–4.

Katebzadeh, M. S., Costa, P., and Grot, B. (2023). Saba: rethinking datacenter network allocation from application’s perspective. In Proc. of the Eighteenth European Conference on Computer Systems, pages 623–638.

Montazeri, B., Li, Y., Alizadeh, M., and Ousterhout, J. (2018). Homa: a receiver-driven low-latency transport protocol using network priorities. In Proc. of the 2018 Conference of the ACM Special Interest Group on Data Communication, pages 221–235.

Prasopoulos, K., Kosta, R., Bugnion, E., and Kogias, M. (2025). SIRD: A sender-informed, receiver-driven datacenter transport protocol. In 22nd USENIX Symposium on Networked Systems Design and Implementation (NSDI 25), pages 451–471.

Reed, D., Gannon, D., and Dongarra, J. (2022). Reinventing high performance computing: challenges and opportunities. arXiv preprint arXiv:2203.02544.

Saltzer, J. H., Reed, D. P., and Clark, D. D. (1984). End-to-end arguments in system design. ACM Transactions on Computer Systems (TOCS), 2(4):277–288.

Tessler, C., Shpigelman, Y., Dalal, G., Mandelbaum, A., Haritan Kazakov, D., Fuhrer, B., Chechik, G., and Mannor, S. (2022). Reinforcement learning for datacenter congestion control. ACM SIGMETRICS Performance Evaluation Review, 49(2):43–46.

Wang, H., Lu, Y., and Wang, Z. (2026). GFCC: a global fast congestion control mechanism based on software-defined networking. IEEE Transactions on Networking, 34:2746–2761.