Application-Aware Aggregation and Traffic Engineering in a Converged Packet-Circuit Network

Das, Saurav; Yiakoumis, Yiannis; Parulkar, Guru; McKeown, Nick; Singh, Preeti; Getachew, Daniel; Desai, Premal Dinesh

doi:10.1364/nfoec.2011.nthd3

Cited by 58 publications

(30 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We prototyped all of these applications on our prototype in fewer than 5000 lines of code, which is at least two orders of magnitude lesser code than an equivalent implementation using current industry-standard solutions [10,22]. Our code is not production ready, mainly because NOX is designed to run on a single server.…”

Section: B Evaluation and Discussionmentioning

confidence: 99%

Rethinking IP Core Networks

Das¹,

Parulkar²,

McKeown³

2013

J. Opt. Commun. Netw.

Self Cite

View full text Add to dashboard Cite

Abstract-The Internet core today is completely based on IP routers. Circuits are only used to provide static point-topoint optical links between routers. As others have recognized, current practice makes it hard to take advantage of very high-speed, low-power optical circuit switches in the core. Despite many proposals to mix packet switching with dynamic circuit switching, none have taken hold. In these times of rapidly increasing traffic, and rapidly decreasing profits for ISPs, it is time to rethink how backbones are built. In this paper, we propose to replace the largest backbone routers with much cheaper hybrid packet-optical switches in a fully meshed IP core. We take advantage of new control planes for software-defined networks to control the packet and circuit networks from a single vantage point. To demonstrate the enormous potential cost savings, we present a detailed analysis of the capital expenditure and show how our approach offers 60% cost savings for typical backbone operators.

show abstract

Section: B Evaluation and Discussionmentioning

confidence: 99%

Rethinking IP Core Networks

Das¹,

Parulkar²,

McKeown³

2013

J. Opt. Commun. Netw.

Self Cite

View full text Add to dashboard Cite

show abstract

“…The work which is most similar to the current one is presented in [6]. It is argued there that a dynamic flow aggregation mechanism is needed in order to have a more manageable number of OF rules in the core switches.…”

Section: Literature and Related Workmentioning

confidence: 91%

“…It is argued there that a dynamic flow aggregation mechanism is needed in order to have a more manageable number of OF rules in the core switches. While the solution proposed in [6] is not described in detail, the criteria used for the aggregation are similar to the ones presented in the current paper: path similarity and similarity of traffic types. The transport protocol and port numbers are used to distinguish between different traffic types.…”

Section: Literature and Related Workmentioning

confidence: 99%

“…This decision is taken by a different entity when marking the packets with the DSCP codes. Another distinction from [6] is that the current paper provides a quantitative analysis of the aggregation service for various key topologies. This analysis gives a precise indication about the degree of savings that the flow aggregation service produces in the flow tables.…”

Section: Literature and Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Dynamic aggregation of traffic flows in SDN: Applied to backhaul networks

Mimidis¹,

Caba²,

Soler³

2016

2016 IEEE NetSoft Conference and Workshops (NetSoft)

View full text Add to dashboard Cite

Abstract-A challenge in the adoption of the OpenFlow (OF)-based SDN paradigm is related to the limited number of OF rules supported by the network devices. The technology used to implement the OF rules is TCAM, which is expensive and power demanding. Due to this, the network devices are either very costly or they can support a very limited number of OF rules. One way to cope with this limitation, is to perform the same logic but with fewer OF rules in the devices. As a demonstration of this operational strategy, the current paper proposes a service for traffic flow aggregation which reduces the number of OF rules needed in the network devices, without impacting the control plane logic. The proposed traffic flow aggregation service is tested on a set of topologies specific to the backhaul network, since they aggregate a large amount of traffic flows. The results illustrate significant reductions in the number of OF rules in network devices, thus a lower demand on TCAM capacity.

show abstract

“…For example, Greenberg et al [8] have proposed the use of programmable commodity switches to reduce network cost and enhance performance through the use of Valiant Load Balancing, Al-Fares et al [9] propose using commodity Ethernet switches to support the full aggregate bandwidth of large clusters, Guo et al [10] use servers as nodes in the interconnect, Das et al [11] explore the use of OpenFlow to control routing according to application need and Webb et al [12] propose to isolate applications and use different routing mechanisms for them in fat-tree based data-centers. The flow of big data traffic in data center networks has also been studied.…”

Section: Related Workmentioning

confidence: 99%

Network Topology Optimization for Data Aggregation

Das

Sahni

2014

2014 14th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing

View full text Add to dashboard Cite

Abstract-In this paper, we show that the problem of configuring the topology of a data center network to optimize data aggregation is NP-hard even when the number of aggregators is 1. Further, the approximation ratio of the algorithm proposed by Wang, Ng, and Shaikh [3] for the case of a single aggregator is (k+1)/2, where k is the degree of ToR (top-of-rack) switches and this algorithm also exhibits an anomalous behavior-increase in the switch degree may result in an increase in the aggregation time. By comparison, if topology configuration is done using the longest processing time (LPT) scheduling rule, the approximation ratio is (4/3-1/(3k)). We show that for every instance of the single aggregator topology configuration problem, the time required to aggregate using the LPT configuration is no more than that using the Wang et al. rule. By coupling the LPT rule with the rule of Wang et al., we achieve a better throughput as promised by LPT and at the same time reduce the total network traffic. Experimental results show that the LPT rule reduces aggregation time by up to 90% compared to the Wang et al. rule. The reduction in aggregation time afforded by a known improvement, COMBINE, of LPT relative to Wang et al. is up to 90.5%. More interestingly, when either of the LPT rule or COMBINE is augmented with the Wang et al. rule, total network traffic is reduced by up to 90% relative to using LPT and COMBINE with chains.

show abstract

Application-Aware Aggregation and Traffic Engineering in a Converged Packet-Circuit Network

Cited by 58 publications

References 3 publications

Rethinking IP Core Networks

Rethinking IP Core Networks

Dynamic aggregation of traffic flows in SDN: Applied to backhaul networks

Network Topology Optimization for Data Aggregation

Contact Info

Product

Resources

About