Network functions virtualization requires automatic deployment and scaling of components. This raises the question of where to place instances of a function, for instance in the OpenStack cloud system. Data plane functions can forward large amounts of traffic. In this case, network-aware placement can avoid an inefficient use of host bandwidth, and a chain of functions can benefit from co-locating instances on a host. However, a practical challenge is that the bandwidth utilization or traffic demand matrix is not always known before the deployment of an instance. A promising remedy is to leverage existing Software Defined Networking (SDN) policies to derive connectivity weights between components. In this paper, we present this novel solution to the online instance placement problem. We have developed an extension of the OpenStack scheduler that uses SDN forwarding policies to rank potential hosts. For a given type of virtual machine, the corresponding forwarding policies can be retrieved from an SDN controller prior to the placement decision. Our prototype identifies potential communication peers and weighs the forwarding rules to prefer hosts that already run communication peers. We present heuristics for such weighing, and we also discuss limitations of the approach. A testbed implementation proofs that even in a simple example our solution can double the service chain throughput.

Cloud computing systems require a placement logic that decides where to allocate resources. In state-of-the-art platforms such as OpenStack, this scheduler takes into account multiple constraints when starting a new instance, including in particular the required computational and memory resources. However, this scheduling mechanism typically neither considers network requirements of Virtual Machines nor the networking resources that are actually available. In this paper we present an extension of the OpenStack scheduler that enables a network-aware placement of instances by taking into account bandwidth constraints to and from nodes. Our solution keeps track of host-local network resource allocation, and it can be combined with bandwidth enforcement mechanisms such as rate limiting. We present a prototype that requires only very few changes in the OpenStack open source software. Testbed measurement results demonstrate the benefit of our solution compared to the OpenStack default approach.

Topology managers expose network information to applications to improve application-level resource management. An example for various ongoing standardization activities is Application-Layer Traffic Optimization (ALTO). Due to privacy and security constraints, exposed network information has to be filtered according to policies. As part of such policies, distance information can be abstracted by presenting coarser distances over pairs of clustered nodes rather than the precise distances between all node pairs. We refer to this process as distance sparsification. The contribution of this paper is two-fold, the first being a new policy system to enforce abstraction. The second and the main contribution addresses the algorithmic challenge. For the latter, we consider two types of distance sparsification algorithms. The first variant takes as input a matrix of pairwise distances. The sparsification algorithm produces a smaller distance matrix by clustering the nodes into clusters. The second variant instead collapses an edge-weighted graph. We measure the performance of the algorithms by the accuracy of the resulting sparsified distances, and we show that matrix sparsification outperforms graph sparsification. We further observe the trade-off between the accuracy and the size of the sparsified representation. In addition, we also extend our algorithms to handle labeled data, i. e., abstraction policies explicitly mark a number of destinations as reference points. Such additional information can improve the distance sparsification.

The ability to extract topology information from the network is important for many applications and enables more informed resource selection. The challenge for topology exposure is to provide a compact representation that is sufficiently accurate and complies to topology hiding policies. This paper presents a topology abstraction system that can expose large-scale service provider network maps with an adjustable level-of-detail. Our system uses graph sparsification algorithms to reduce the complexity of routing topologies. Our numerical results reveal that the size of maps can be reduced by one order of magnitude or more while the result still enables reasonable traffic optimization inside applications. A proof-of-concept implementation gathers network management system data and exposes abstract maps through the Application-Layer Traffic Optimization (ALTO) protocol.

Multipath transport protocols such as Multipath TCP can concurrently use several subflows to transmit a TCP flow over potentially different paths. Since more than one subflow is used, an efficient multipath scheduling algorithm is needed at the sender. The objective of the scheduler is to identify the subflow over which the current data packet should be sent. This paper compares the most important types of schedulers for multipath transfers. We model their performance analytically and derive key metrics, most notably the resulting end-to-end delay over heterogeneous paths. Our results show that a scheduler minimizing the packet delivery delay yields the best overall performance, but it is complex to realize. An alternative scheduler based on the sender queue size is simpler and has sufficient performance for relatively small asymmetry between the multiple paths. Our model results are confirmed by measurements with a real multipath transport protocol.

The Software Defined Network (SDN) model depends on abstractions to separate the control plane from the packet forwarding plane. Applications can interact with the control plane to receive a global network view, upon which they can operate. By having access to network topology information, applications can optimize decisions related to service rendezvous, service fulfillment, service placement and service removal. The network is in the best position to provide guidance to a broad class of applications, including peer-to-peer systems, Content Distribution Network (CDN), and datacenter applications. In all these use cases proximity matters as peers need to rendezvous with other peers and users need to rendezvous with the best cloud application or best CDN server. We maintain that a solution for such a rendezvous problem should be an intrinsic component of the emerging SDN model. A specific instance of a protocol that abstracts network topology is the Application Layer Traffic Optimization (ALTO) protocol. ALTO provides applications an abstract view of the network and thus enables applications to leverage a network without exposing the network provider's internal details or policies. We argue that ALTO provides a clean, mature, standards-based and powerful abstraction, which can be used by SDNs today to obtain network information for solving the rendezvous problem.

More and more services are provided by large data centers with a potentially very large number of physical or virtual hosts. As the number of hosted services and service consumers increases, also the number of hosts inside a data center raises to cope with the increasing end-user demand. Current data center networks are usually based on Ethernet and mechanisms like load balancing or redundancy between data centers require a transparent connection of these Ethernet networks over a Wide Area Network (WAN). Due to the large number of hosts, these interconnected data center networks face scalability problems on different protocol layers. One such issue, which is currently discussed within the IETF, is the scalability of the link layer Address Resolution Protocol (ARP). This paper studies the control traffic caused by address resolution for interconnected data centers. We develop an analytical model for the ARP traffic between data center locations that takes into account the number of hosts and connected sites. This model can then be used to quantify the ARP traffic for a data center interconnect solution. As an example, we apply our model to Virtual Private LAN Services (VPLS). In addition, we study how an ARP proxy can improve the overall scalability, and we show that a proxy significantly reduces the ARP traffic at VPLS switches.

We consider a system of compute and storage resources geographically distributed over a large number of locations connected via a wide-area network. By distributing the resources, latency to users can be decreased, bandwidth costs reduced and availablility increased. The challenge is to distribute services with varying characteristics among the data centers optimally. Some services are very latency sensitive, others need vast amounts of storage, and yet others are computationally complex but do not require hard deadlines on execution. We propose efficient algorithms for the placement of services to get the maximum benefit from a distributed cloud systems. The algorithms need input on the status of the network, compute resources and data resources, which are matched to application requirements.

Multipath transport protocols enable the concurrent use of different paths and resource pooling of their capacity. Ongoing research and standardization activities extend the Transmission Control Protocol (TCP) towards a Multipath TCP solution. The objective is to aggregate multiple subflows between two endpoints into a single session that can be accessed by an application like a single TCP connection. This paper presents and evaluates Multi-Connection TCP (MCTCP) as a new multipath transport solution. MCTCP consists of a shim layer on top of several TCP connections and encodes control information, as far as possible, in their payload. In this paper, we first motivate MCTCP's design and compare it to a multipath transport protocol that uses TCP option encoding only. Second, we show that MCTCP can be implemented in the Linux stack with very few kernel modifications. We also demonstrate that MCTCP can be combined with a coupled multipath congestion control mechanism. Measurement results prove that MCTCP is an efficient and robust protocol that can dynamicall.y pool the bandwidth of several paths.

We recently witness new services that can afford some delay until data transmission starts, but then benefit from a very large available bandwidth. A popular example is the migration of virtual machines between different sites of a geographically dispersed service provider. In this paper, we propose a subscription model for time-scheduled data transfers. Transmission requests are served consecutively, giving the flows access to the physical bandwidth. This is in contrast to today's Internet where flows are served in parallel so that they compete for the available bandwidth. We present the architecture to enable such data beams. Furthermore, we model and analyze the performance under different conditions and compare it with concurrent transmission.

The IP Multimedia Subsystem (IMS) serves as universal platform for fast and standardized creation of mobile services. Typical deployment models for the IMS favor centralized session control and application servers. Furthermore, they rely on sophisticated border elements for, e.g., policy enforcement. One step towards a simpler architecture is the integration of call control, application server, and media functions into border elements. This leads to a distributed IMS architecture with equal processing nodes, which drastically reduces system complexity and scales on functional module instantiation basis. However, the distributed nature of the architecture reduces the statistical gain. In this work, we study a suitable load sharing concept that counters or even over-compensates this problem. In principle, this mechanism realizes one large virtual central server with small control and message overhead. Our simulative evaluation shows the feasibility and performance of our approach.

Quick-Start is an experimental extension of the Transmission Control Protocol (TCP) that allows to speed up best effort data transfers. With Quick-Start, TCP hosts can request permission from the routers along a network path to send at a higher rate than allowed by the default TCP congestion control. The explicit router feedback avoids the time-consuming capacity probing by the TCP Slow-Start and is therefore particularly beneficial for underutilized paths with a high bandwidth-delay product. In this paper, the performance of the Quick-Start TCP extension is analyzed and the impact of router admission control strategies is studied. The main contribution is an analytical model that quantifies the improvement compared to default TCP Slow-Start. The model is validated by simulation results and by initial measurements with a Quick-Start implementation in the Linux operating system. Our results confirm that Quick-Start can significantly reduce the completion times of mid-sized data transfers.

Cellular networks are more and more complemented by wireless LAN hotspots. However, when users are mobile, the small coverage area of hotspots is a challenge for the access network discovery and selection. Network selection mechanisms can be improved if, in addition to terminal-based measurements, hotspot information can be retrieved from an information service. In this context, the coverage area is one of the most important characteristics of hotspots. This paper studies algorithms for the precise estimation of hotspot coverage areas based on field strength measurements performed by individual terminals. Different approaches for interpolation, vectorization and complexity reduction of measurement data are presented and evaluated in different scenarios by means of simulation. As a result, guidelines for the optimal parametrization of these algorithms are presented.

Many signaling protocols in IP networks need a protection against message loss, but they do not require a strict in-sequence data delivery. Since TCP provides reliable in-order transport, end-to-end delays may be unnecessarily increased due to head-of-line blocking. An alternative transport protocol is SCTP, which is optimized for signaling applications and provides mechanisms for reliable, partial-ordered or unordered message delivery. In this paper, we quantify the impact of head-of-line blocking on the response time of transaction-based signaling applications. In order to mitigate this problem, we compare different solutions based on TCP and SCTP. Both a new analytical model and measurements on state-of-the-art operating systems show to which extend SCTP can improve transport delays by leveraging transmission over multiple parallel streams or using unordered data delivery. Our analysis reveals that using one or multiple parallel TCP connections can result in much higher end-to-end delays, even for moderate packet loss probabilities. We also observe significant differences in the TCP performance of different operating systems.

Mobility models are widely used in simulation-based performance analyses of mobile networks. However, there is a trade-off between simplicity and realistic movement patterns. Synthetic models like the random waypoint and random direction model are simple to implement, but only provide unrealistic simple user sojourn densities and traffic flows. In contrast, graph and trip-based mobility models are complex to parameterize and their results are difficult to compare. In this paper, we propose the location-dependent parameterization of the random direction model to fill this gap. This model extension allows to setup non-homogeneous mobility scenarios, in particular based on real-world traces, while it still belongs to the class of synthetic random walk mobility models. We show that the location-dependent parametrization can accurately model arbitrary mobility patterns with very limited implementation complexity.

MOTIVATION: The Protein Data Bank currently contains more than 4700 protein coordinate sets. It is often desirable to make a selection from these files based on a criterion like R-factor, experimental method, length of the amino acid sequence, or the number of homologous sequences in SWISSPROT. Doing this using the distributed form of the Protein Data Bank can be a tedious task, because (1) this requires reading one file for every single entry, and (2) not all of the information is present in a consistent computer readable way in all of the entries. RESULTS: The PDBFINDER database provides an easy to interpret file containing summary information about all Protein Data Bank files. Summary information from the DSSP (Definition of Secondary Structure of Proteins) and HSSP (Homology derived Secondary Structure of Proteins) databases is also included. Furthermore, where essential data were missing from the Protein Data Bank file, this information has been retrieved from the original literature. AVAILABILITY: The latest version of the PDBFINDER database can be downloaded by anonymous ftp from swift.embl-heidelberg.de, directory:/pdbfinder. CONTACT: E-mail address [email protected].