RDNA Balance: Load Balancing by Elephant Flow Isolation in Data Centers with Source Routing

• Rodolfo V. Valentim Universidade Federal do Espirito Santo
• Cristina K. Dominicini Universidade Federal do Espirito Santo / Instituto Federal do Espírito Santo
• Diego R. Mafioletti Universidade Federal do Espirito Santo / Instituto Federal do Espírito Santo
• Rodolfo S. Villaca Universidade Federal do Espirito Santo
• Moises R. N. Ribeiro Universidade Federal do Espirito Santo
• Magnos Martinello Universidade Federal do Espírito Santo

DOI: https://doi.org/10.5753/sbrc.2019.7418

Abstract

Data center networks, which need to dynamically service a large number of flows with different service requirements, require load balancing mechanisms. However, traditional load balancing approaches do not allow full utilization of network resources in a simple, programmable and scalable manner. In this context, this paper proposes RDNA Balance that explores elephant stream isolation balancing and source routing, supported by the core network, and edge sorting operations using existing features in the OpenFlow protocol. The results show that this approach is capable of providing simple, scalable, and programmable load balancing.

Keywords: Data Center Networks, Load Balancing, OpenFlow

References

Al-Fares, M. et al. (2010). Hedera: Dynamic flow scheduling for data center networks. In Proceedings of the 7th USENIX NSDI, NSDI’10, pages 19–19, CA, USA. USENIX.

Alizadeh, M. et al. (2014). CONGA: Distributed Congestion-Aware Load Balancing for Datacenters. ACM SIGCOMM Computer Communication Review, 44(4):503–514.

Benson, T. et al. (2010). Understanding data center traffic characteristics. SIGCOMM Comput. Commun. Rev., 40(1):92–99.

Bolla, R. and Bruschi, R. (2006). Rfc 2544 performance evaluation and internal measurements for a linux based open router. In High Performance Switching and Routing, 2006 Workshop on, pages 6–pp. IEEE.

Cai, Y. et al. (2012). RFC 6754 - protocol independent multicast equal-cost multipath (ECMP) redirect. Technical report.

He, K. et al. (2015). Presto: Edge-based Load Balancing for Fast DCN. Sigcomm 2015, pages 465–478.

Jin, X. et al. (2016). Your data center switch is trying too hard. In Proceedings of the Symposium on SDN Research, SOSR ’16, pages 12:1–12:6, NY, USA. ACM.

Jyothi, S. A. et al. (2015). Towards a flexible data center fabric with source routing. In Proceedings of the 1st ACM SIGCOMM Symposium on Software Defined Networking Research, page 10. ACM.

Kandula, S. et al. (2009). The nature of data center traffic. In Proceedings of the 9th ACM SIGCOMM conference on IMC ’09, page 202, New York, New York, USA. ACM.

Katta, N. et al. (2015). Ravana: Controller fault-tolerance in software-defined networking. In Proceedings of the 1st ACM SIGCOMM symposium on software defined networking research, page 4. ACM.

Katta, N. et al. (2016). Hula: Scalable load balancing using programmable data planes. In Proceedings of the Symposium on SDN Research, SOSR ’16, pages 10:1–10:12, NY, USA. ACM.

Liberato, A. et al. (2018). RDNA: Residue-Defined Networking Architecture Enabling Ultra-Reliable Low-Latency Datacenters. IEEE TNSM, 4537(c):1–1.

Martinello, M. et al. (2014). Keyflow: A prototype for evolving SDN toward core network fabrics. Network, IEEE, 28:12–19.

Popoviciu, C., Hamza, A., Van de Velde, G., and Dugatkin, D. (2008). RFC 5180 - IPv6 benchmarking methodology for network interconnect devices. Technical report.

Rasley, J. et al. (2014). Planck: Millisecond-scale Monitoring and Control for Commodity Networks Jeff. In Proceedings of the 2014 ACM conference on SIGCOMM - SIGCOMM ’14, pages 407–418, New York, New York, USA. ACM Press.