Sujata Banerjee

DevoFlow: scaling flow management for high-performance networks

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 switchs 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 OpenFlows current design cannot meet the needs of high-performance 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.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 switchs 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 OpenFlows current design cannot meet the needs of high-performance 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.

A Power Benchmarking Framework for Network Devices

Energy efficiency is becoming increasingly important in the operation of networking infrastructure, especially in enterprise and data center networks. Researchers have proposed several strategies for energy management of networking devices. However, we need a comprehensive characterization of power consumption by a variety of switches and routers to accurately quantify the savings from the various power savings schemes. In this paper, we first describe the hurdles in network power instrumentation and present a power measurement study of a variety of networking gear such as hubs, edge switches, core switches, routers and wireless access points in both stand-alone mode and a production data center. We build and describe a benchmarking suite that will allow users to measure and compare the power consumed for a large set of common configurations at any switch or router of their choice. We also propose a network energy proportionality index, which is an easily measurable metric, to compare power consumption behaviors of multiple devices.

Distributed mechanisms for quality of service in wireless LANs

Wireless local area networks are gaining popularity at an unprecedented rate, at home, at work, and in public hot spot locations. As these networks become ubiquitous and an integral part of the infrastructure, they will increasingly be used for multimedia applications. There is limited QoS support in WLANs, which will become an impediment in deploying multimedia applications. We present a tutorial on QoS support in IEEE 802.11 WLANs with a focus on the distributed MAC protocol of 802.11. Most QoS support mechanisms proposed for 802.11 use well-known techniques such as priority assignment and fair scheduling, and map QoS metrics into some existing 802.11 MAC parameter, thereby avoiding a redesign of the MAC protocol. We provide a taxonomy of the mechanisms and describe the essential concepts, problems, and advantages of each mechanism. From our study, we conclude that choosing the right set of MAC parameters and the QoS mechanism itself to provide predictable QoS in 802.11 networks is still an open problem.

ElasticSwitch: practical work-conserving bandwidth guarantees for cloud computing

While cloud computing providers offer guaranteed allocations for resources such as CPU and memory, they do not offer any guarantees for network resources. The lack of network guarantees prevents tenants from predicting lower bounds on the performance of their applications. The research community has recognized this limitation but, unfortunately, prior solutions have significant limitations: either they are inefficient, because they are not work-conserving, or they are impractical, because they require expensive switch support or congestion-free network cores. In this paper, we propose ElasticSwitch, an efficient and practical approach for providing bandwidth guarantees. ElasticSwitch is efficient because it utilizes the spare bandwidth from unreserved capacity or underutilized reservations. ElasticSwitch is practical because it can be fully implemented in hypervisors, without requiring a specific topology or any support from switches. Because hypervisors operate mostly independently, there is no need for complex coordination between them or with a central controller. Our experiments, with a prototype implementation on a 100-server testbed, demonstrate that ElasticSwitch provides bandwidth guarantees and is work-conserving, even in challenging situations.

S 3 : a scalable sensing service for monitoring large networked systems

Efficiently operating and managing large scale distributed and federated systems is an extremely challenging problem. Current solutions are a combination of centralized management and significant over-provisioning of the infrastructure. With the explosion of new resource-intensive media applications and services, over provisioning of the infrastructure is no longer a viable option. Timely and accurate knowledge of the global environment (particularly the highly dynamic network path properties) is necessary for management of performance SLAs, just-in-time resource provisioning, near-optimal dynamic service placement and reuse, construction of network service overlays, and fast detection of failures and malicious attacks. Further, different applications require information about different aspects of the environment at different timescales. We propose S3, a Scalable Sensing Service, that achieves the above requirements and enables personalized sensing of the environment as dictated by applications.

DevoFlow: cost-effective flow management for high performance enterprise networks

The OpenFlow framework enables flow-level control over Ethernet switching, as well as centralized visibility of the flows in the network. OpenFlows coupling of these features comes with costs, however: the distributed-system costs of involving the OpenFlow controller on flow setups, and the switch-implementation costs of involving the switchs control plane too often. In this paper, we analyze the overheads, and we propose DevoFlow, a modification of the OpenFlow model in which we try to gently break the coupling between centralized control and centralized visibility, in a way that maintains a useful amount of visibility without imposing unnecessary costs.

NodeWiz: peer-to-peer resource discovery for grids

Efficient resource discovery based on dynamic attributes such as CPU utilization and available bandwidth is a crucial problem in the deployment of computing grids. Existing solutions are either centralized or unable to answer advanced resource queries (e.g., range queries) efficiently. We present the design of NodeWiz, a grid information service (CIS) that allows multi-attribute range queries to be performed efficiently in a distributed manner. This is obtained by aggregating the directory services of individual organizations in a peer-to-peer information service.

Measuring bandwidth between planetlab nodes

With the lack of end-to-end QoS guarantees on existing networks, applications that require certain performance levels resort to periodic measurements of network paths. Typical metrics of interest are latency, bandwidth and loss rates. While the latency metric has been the focus of many research studies, the bandwidth metric has received comparatively little attention. In this paper, we report our bandwidth measurements between PlanetLab nodes and analyze various trends and insights from the data. For this work, we assessed the capabilities of several existing bandwidth measurement tools and describe the difficulties in choosing suitable tools as well as using them on PlanetLab.

Optimization of indoor wireless communication network layouts

Radio Frequency Data Communications (RFDC)technology is rapidly becoming a critical component of many traditional industrial engineering functions including materials tracking, inventory control, warehousing, order processing, shipping and database management. As a means of moving information, RFDC has many attractive features, such as speed, accuracy, reliability, convenience and low operating costs. When implementing RFDC systems a major problem is to quickly and efficiently determine the locations where transceivers should be placed so that effective radio communication can take place. The research described in this paper addresses this issue by developing a computerized layout simulation system that incorporates heuristic optimization methods to solve the placement problem. The effectiveness of this unique automated layout methodology is demonstrated by comparing it with the current method of utilizing manual site surveys, as well as with other placement methods. The methodology and solutions are validated by field-testing at actual facilities.

Application-driven bandwidth guarantees in datacenters

Providing bandwidth guarantees to specific applications is becoming increasingly important as applications compete for shared cloud network resources. We present CloudMirror, a solution that provides bandwidth guarantees to cloud applications based on a new network abstraction and workload placement algorithm. An effective network abstraction should enable applications to easily and accurately specify their requirements, while simultaneously enabling the infrastructure to provision resources efficiently for deployed applications. Prior research has approached the bandwidth guarantee specification by using abstractions that resemble physical network topologies. We present a contrasting approach of deriving a network abstraction based on application communication structure, called Tenant Application Graph or TAG. CloudMirror also incorporates a new workload placement algorithm that efficiently meets bandwidth requirements specified by TAGs while factoring in high availability considerations. Extensive simulations using real application traces and datacenter topologies show that CloudMirror can handle 40% more bandwidth demand than the state of the art (e.g., the Oktopus system), while improving high availability from 20% to 70%.