Library

Description: Executing applications in the Grid often requires access to multiple geographically distributed resources. In a Grid environment, these resources belong to different administrative domains, each employing its own scheduling policy. That is, at which time an activity (e.g., compute job, data transfer) is started, is decided by the resource's local management system. In such an environment, the coordinated execution of distributed applications requires guarantees on the quality of service (QoS) of the needed resources. Reserving resources in advance is an accepted means to obtain QoS guarantees from a single provider. The challenge, however, is to coordinate advance reservations of multiple resources. This work presents a system architecture and mechanisms to coordinate multiple advance reservations -- called co-reservations -- for delivering QoS guarantees to complex applications. We formally define the co-reservation problem as an optimization problem. The presented model supports three dimensions of freedom: the start time, the duration and the service level of a reservation. Requests and resources are described in a simple language. After matching the static properties and requirements of either side in a mapping, the reservation mechanism probes information about the future status of the resources. The versatile design of the probing step allows the efficient processing of requests, but also lets the resources express their preferences among the myriads of reservation candidates. Next, the best mapping is found through an implementation of the formal co-reservation model. Then, the mapping has to be secured, i.e., resources need to be allocated to a co-reservation candidate with all-or-nothing semantics. We study several goal-driven sequential and concurrent allocation mechanisms and define schemes for handling allocation failures. Finally, we introduce the concept of virtual resources for seamlessly embedding co-reservations into Grid resource management.

Description: Die Ausführung von Anwendungen erfordert oft mehrere, geographisch verteilte Ressourcen. In Grid-Umgebungen gehören diese Ressourcen zu verschiedenen administrativen Organisationen, wobei jede ihre eigene Schedulingregeln verwendet. Das bedeutet, zu welcher Zeit eine Aktivität gestartet wird (z.B. ein Rechenjob), wird vom lokalen Ressourcenmanagementsystem entschieden. Die koordinierte Ausführung von verteilten Anwendungen erfordert Dienstgütegarantien für die benötigten Ressourcen. Das Reservieren von Ressourcen im Voraus ist ein Mittel, um Dienstgütegarantien von einem einzelnen Ressourcenanbieter zu erhalten. Die Herausforderung in dieser Arbeit ist, Vorausreservierungen von mehreren Ressourcen zu koordinieren. Es wird ein System für die Koordinierung mehrerer Vorausreservierungen -- Co-Reservierungen genannt -- für die Bereitstellung von Dienstgütegarantien vorgestellt. Wir definieren das Co-Reservierungsproblem als Optimierungsproblem. Das vorgestellte Modell unterstützt drei Freiheitsgrade: die Startzeit, die Dauer und die Dienstgüte einer Reservierung. Anfragen und Ressourcen werden in einer einfachen Sprache beschrieben. Nachdem statische Eigenschaften und Anforderungen beider Seiten überprüft wurden, ermittelt der Reservierungsmechanismus Informationen über den zukünftigen Zustand der Ressourcen. Dieser Schritt ist so allgemein gehalten, daß er sowohl ein effizientes Bearbeiten der Anfragen erlaubt als auch den Ressourcen ermöglicht ihre Präferenzen auszudrücken. Im Anschluss wird die optimale Zuweisung von Anfragen zu Ressourcen ermittelt. Im letzten Schritt muss diese Zuweisung umgesetzt werden, d.h., entweder alle oder keine Ressource wird allokiert. Es werden mehrere sequentielle und parallele Allokationsverfahren vorgestellt sowie deren Auswirkung auf verschiedene Metriken untersucht. Die Einbettung von Co-Reservierungen in das Grid-Ressourcenmanagement wird anhand des Konzeptes der virtuellen Ressource dargestellt.

Description: One of the biggest impacts on the performance of a Distributed Hash Table (DHT), once established, is its ability to balance load among its nodes. DHTs supporting range queries for example suffer from a potentially huge skew in the distribution of their items since techniques such as consistent hashing can not be applied. Thus explicit load balancing schemes need to be deployed. Several such schemes have been developed and are part of recent research, most of them using only information locally available in order to scale to arbitrary systems. Gossiping techniques however allow the retrieval of fairly good estimates of global information with low overhead. Such information can then be added to existing load balancing algorithms that can use the additional knowledge to improve their performance. Within this thesis several schemes are developed that use global information like the average load and the standard deviation of the load among the nodes to primarily reduce the number of items an algorithm moves to achieve a certain balance. Two novel load balancing algorithms have then been equipped with implementations of those schemes and have been simulated on several scenarios. Most of these variants show better balance results and move far less items than the algorithms they are based on. The best of the developed algorithms achieves a 15-30% better balance and moves only about 50-70% of the number of items its underlying algorithm moves. This variation is also very robust to erroneous estimates and scales linearly with the system size and system load. Further experiments with self-tuning algorithms that set an algorithm’s parameter according to the system’s state show that even more improvements can be gained if additionally applied. Such a variant based on the algorithm described by Karger and Ruhl shows the same balance improvements of 15-30% as the variant above but reduces the number of item movements further to 40-65%.

Description: We present a framework for transactional data access on data stored in a DHT. It allows to atomically read and write items and to run distributed transactions consisting of a sequence of read and write operations on the items. Items are symmetrically replicated in order to achieve durability of data stored in the SON. To provide availability of items despite the unavailability of some replicas, operations on items are quorum-based. They make progress as long as a majority of replicas can be accessed. Our framework processes transactions optimistically with an atomic commit protocol that is based on Paxos atomic commit. We present algorithms for the whole framework with an event based notation. Additionally we discuss the problem of lookup inconsistencies and its implications on the one-copy serializability property of the transaction processing in our framework.

Description: When planning teams for projects with specific goals, employees of a company have to group together so well, that all necessary knowledge for conquering the project’s challenges are met within the member’s skills. A tool that facilitates semantic web technologies can support the team recruiter, who is responsible for chosing the members of the team, in terms of finding the most efficient combinations of the company’s employees based on their expertises.

Description: Key/value stores which are built on structured overlay networks often lack support for atomic transactions and strong data consistency among replicas. This is unfortunate, because consistency guarantees and transactions would allow a wide range of additional application domains to benefit from the inherent scalability and fault-tolerance of DHTs. The Scalaris key/value store supports strong data consistency and atomic transactions. It uses an enhanced Paxos Commit protocol with only four communication steps rather than six. This improvement was possible by exploiting information from the replica distribution in the DHT. Scalaris enables implementation of more reliable and scalable infrastructure for collaborative Web services that require strong consistency and atomic changes across multiple items.

Description: The Cooperative Library Network Berlin-Brandenburg (KOBV), Germany, addresses the problem of how to integrate resources found outside the library and library holdings into a single discovery tool. It presents a solution that uses open source technology to develop a next-generation catalog interface called the Library Search Engine. This pilot project was launched in 2007 with the library of Albert Einstein Science Park, Potsdam. The idea was to design and develop a fast and convenient search tool, integrating local holdings (books, journals, journal articles) as well as relevant scientific subject information such as open access publications and bibliographies.

Description: Asymptotic analyses of the three dimensional compressible flow equations coupled with transport equations for the mixing ratios of water vapour, cloud water and rain water are described. We obtain reduced systems of equations for two particular regimes of length and time scales: Models for the long time evolution of deep convective columns and for the short time evolution of shallow convective layers. The asymptotic deep convective column model is anelastic, yet the vertical motion is pressure free, i.e., it evolves freely in interaction with buoyancy while the horizontal divergence adjusts to fullfil the anelastic constraint. The perturbation pressure guaranteeing compliance with the horizontal divergence constraint obeys a Poisson-type equation. Surprisingly, the vertical velocity plays an important role in the horizontal dynamics through the Coriolis term. The vertical acceleration in a saturated column is directly determined by the buoyancy induced by potential temperature differences relative to the background stratification. This potential temperature deviation is a conserved quantity. Evaporation is the only important microphysical process in the undersaturated regime. The evaporation rate depends on the saturation deficit and the amount of rain water present and determines the (downward) vertical velocity and the distribution of water vapour. To connect the deep convective column solutions to top and bottom boundary conditions, a different flow regime needs to be accounted for. Within shallow layers whose depth is comparable to the column diameters, adjustment to physical boundary conditions can take place. This is the second regime considered in this report. The shallow convective layer regime is shown to be asymptotically described by Boussinesq-type equations. These equations are closed by evolution equations which show that, in the saturated regime, the distributions of potential temperature and cloud water are determined by a condensation rate that is directly proportional to the vertical velocity. In the undersaturated regime, the potential temperature distribution is determined by the amount of rain present, since the water vapour in this case is shown to be a conserved quantity. In both regimes the distribution of rain water depends on the rain water flux.

Description: By use of asymptotic analysis Carqué et al. [ZIB-Report 08-03] derived an asymptotic column model for deep convective clouds based on the three dimensional compressible flow equations and a bulk microphysics parameterization. In the present study we check the plausibility of the reduced model equations by comparing implications of the model for the scaling of various terms in the governing equations with those extracted from large eddy simulation data based on the computational model UCLA-LES1.1. This code solves an anelastic system of equations with complete droplet based microphysics and LES closures. We observe that the simulation data corroborate the basic assumptions of the asymptotic analysis and the main conclusions implied by the asymptotically reduced model. The code output reflects the scales of space and time: The deep convective clouds show an anisotropic structure where the horizontal scale is considerably narrower than the vertical scale; with a period of about 20 min, from emergence to breakup, the life cycle of one particular deep convective cloud corresponds exactly to the reference time of the reduced model. The characteristic properties of dynamics as predicted by the reduced model are also reflected in the simulation data: The horizontal flow is controlled by the pressure field; the vertical velocity develops freely independent of pressure over the depth of the convective column; the vertical velocity is directly determined by the buoyancy induced by the potential temperature deviation relative to the background stratification. With respect to grid resolution we observe that refining the spatial step size of the equidistant computational grid from 125 m to 62.5 m does not influence the results: Even with the coarser grid the relevant physical phenomena are sufficiently resolved. Somewhat surprisingly, the Coriolis term involving vertical velocity and acting on the horizontal (east-west) velocity component appears at leading order in the asymptotics. Accordingly, we expected to find a nontrivial impact of this Coriolis effect on the horizontal flow velocity components within columns of updrafts. However, switching the term on and off in subsequent simulations did not sizeably affect the results.

Description: A theoretical and numerical small-scale study of the evaporative cooling phenomenon that might appear in the stratocumulus-topped boundary layers is presented. An ideal configuration of a cloud-top mixing layer is considered as defined by two nonturbulent horizontal layers, stably stratified and with buoyancy reversal within a certain range of mixture fractions due to the evaporative cooling. Linear stability analysis of the shear-free configuration is employed to provide a new interpretation of the buoyancy reversal parameter, namely, in terms of a time-scale ratio between the unstable and the stable modes of the system. An incompressible high-order numerical algorithm to perform direct numerical simulation of the configuration is described and two-dimensional simulations of single-mode perturbations are discussed. These simulations confirm the role of the different parameters identified in the linear stability analysis and show that convoluted flow patterns can be generated by the evaporative cooling even for the low levels of buoyancy reversal found in stratocumulus clouds. They also show that there is no enhancement of entrainment of upper layer fluid in the shear-free configuration, and mixing enhancement by the evaporative cooling is restricted to the lower layer.

Description: We study performance and scaling of the Berlin Quantum Chromodynamics Program (BQCD) on the SGI Altix 4700 at Leibniz Supercomputing Centre (LRZ). We employ different communication methods (MPI, MPI with two OpenMP threads per process, as well as the shmem library) and run the MPI version on the two types of nodes of that machine. For comparison with other machines we made performance measurements on an IBM p690 cluster and a Cray XT4.

Description: This work introduces a novel streamline seeding technique based on dual streamlines that are orthogonal to the vector field, instead of tangential. The greedy algorithm presented here produces a net of orthogonal streamlines that is iteratively refined resulting in good domain coverage and a high degree of continuity and uniformity. The algorithm is easy to implement and efficient, and it naturally extends to curved surfaces.

Description: In dieser Arbeit wird eine neue Strategie zur Platzierung von Stromlinien vorgestellt. Hierzu werden zusätzliche duale Stromlinien verwendet, die --im Gegensatz zur üblichen Definition-- orthogonal zum Vektorfeld verlaufen. Der vorgestellte Greedy-Algorithmus berechnet ein Netz aus orthogonalen Stromlinien, welches iterativ verfeinert wird, was zu einer guten Abdeckung der Domäne und einer gleichmäßigen Verteilung der Stromlinien führt. Es handelt sich um einen einfach zu implementierenden und effizienten Algorithmus, der direkt auf gekrümmten Oberflächen anwendbar ist.