DevoFlow

Abstract

OpenFlow is a great concept, but its original design imposes excessive overheads. It can simplify network and traffic management in enterprise and data center environments, because it enables flow-level control over Ethernet switching and provides global visibility of the flows in the network. However, such fine-grained control and visibility comes with costs: the switch-implementation costs of involving the switch's control-plane too often and the distributed-system costs of involving the OpenFlow controller too frequently, both on flow setups and especially for statistics-gathering.In this paper, we analyze these overheads, and show that OpenFlow's current design cannot meet the needs of highperformance networks. We design and evaluate DevoFlow, a modification of the OpenFlow model which gently breaks the coupling between control and global visibility, in a way that maintains a useful amount of visibility without imposing unnecessary costs. We evaluate DevoFlow through simulations, and find that it can load-balance data center traffic as well as fine-grained solutions, without as much overhead: DevoFlow uses 10-53 times fewer flow table entries at an average switch, and uses 10-42 times fewer control messages.

Improvement in Multi-domains Routing Model Based on SDN

“…Frequent control information transfer between switches and controllers is an important factor limiting of network transmission performance. Pre-routing can reduce the generation and transmission of control information to a certain extent [10]. The increase, decrease, and movement of network devices have a certain impact on the centralized management of the network.…”

Tolerating SDN Application Failures with LegoSDN

“…We note that serialization and de-serialization of messages, and the communication protocol overhead introduce additional latency into the control-loop (between the switch and the SDN-Apps). The additional latency, however, is acceptable as introducing the controller into the critical-path (of flow setup or packet processing) already slows down the network by a factor of four [11].…”

H‐SOFT: a heuristic storage space optimisation algorithm for flow table of OpenFlow

“…The authors in [16] proposed ACL compressor to handle one-dimensional and multi-dimensional ACL, which can significantly reduce the number of rules by 50.22% in an ACL while maintaining the same semantics. Curtis et al [17] pointed out that the OpenFlow technology is based on data flow instead of the destination address, resulting in the more entries included in flow table of OpenFlow switches than those in traditional routing table under the same network traffic conditions. In other words, one data flow may correspond to multiple flow entries to achieve the control over different granularity.…”

“…According to the characteristics of flow table in OpenFlow, Curtis et al [17] proposed the algorithm to reduce the number of flow entries by revising the DevoFlow model of OpenFlow, but this algorithm has low extensibility. The authors in [18] presented the DIFANE model which combines the active and passive installation scheme of flow tables to keep traffic in the forwarding plane and reduce the message exchange between the controller and switches, and thus lower the traffic load of the controller.…”