9 results
Now showing1 - 9 of 9
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , VTDL: A notation for data stream processing applicationsThe continuing growth of the Internet of Things (IoT) requires established stream processing engines (SPEs) to cope with new challenges, like the geographic distribution of IoT sensors and clouds hosting the SPEs. These challenges obligate SPEs to support distributed stream processing across different geographic locations which also require a new approach on how data stream processing topologies are defined. In this paper, we identify required features for next-generation SPEs and introduce the Vienna Topology Description Language (VTDL). This language is specifically designed to address challenges for next-generation SPEs and proposes several novel aspects compared to existing topology description concepts. To assess not only the feasibility but also the reduced management overhead due to the VTDL, we evaluate the VTDL within the VISP stream processing ecosystem and show that the usage of the VTDL approach results in a management time reduction of up to 18 times.Publicationtype: Conference Paper40 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Elastic stream processing for the internet of thingsEmerging trends like Big Data and the Internet of Things pose new challenges to established data stream processing engines. Especially, with the advent of the Internet of Things, the data that has to be processed can become very large. Since companies usually aim for cost efficiency, engines need to support resource elasticity to minimize the operational cost while maintaining real-time processing capabilities. In the work at hand, we propose and realize the distributed Platform for Elastic Stream Processing (PESP). An extensive evaluation demonstrates the practical feasibility and efficiency of the system design. The evaluation shows that PESP is able to reduce cost by 20% with minimal effects on the Quality of Service in comparison to an over-provisioning baseline. Compared to an under-provisioning baseline, PESP allows a Quality of Service improvement of 72%.Publicationtype: Conference Paper43 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Nomadic applications traveling in the fogThe emergence of the Internet of Things introduces new challenges like network congestion or data privacy. However, it also provides opportunities, such as computational resources close to data sources, which can be pooled to realize Fogs to run software applications on the edge of the network. To foster this new type of resources, we revisit the concept of mobile agents and evolve them to so-called nomadic applications, which allow addressing vital challenges for the Internet of Things. In this paper, we propose a system design to realize nomadic applications and discuss several challenges that need to be addressed in order to apply them to real-world scenarios.Publicationtype: Conference Paper22 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Cost-efficient enactment of stream processing topologiesThe continuous increase of unbound streaming data poses several challenges to established data stream processing engines. One of the most important challenges is the cost-efficient enactment of stream processing topologies under changing data volume. These data volume pose different loads to stream processing systems whose resource provisioning needs to be continuously updated at runtime. First approaches already allow for resource provisioning on the level of virtual machines (VMs), but this only allows for coarse resource provisioning strategies. Based on current advances and benefits for containerized software systems, we have designed a cost-efficient resource provisioning approach and integrated it into the runtime of the Vienna ecosystem for elastic stream processing. Our resource provisioning approach aims to maximize the resource usage for VMs obtained from cloud providers. This strategy only releases processing capabilities at the end of the VMs minimal leasing duration instead of releasing them eagerly as soon as possible as it is the case for threshold-based approaches. This strategy allows us to improve the service level agreement compliance by up to 25% and a reduction for the operational cost of up to 36%.Publicationtype: Journal Article24 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Optimization of Complex Elastic ProcessesBusiness Process Management is a matter of great importance in different industries and application areas. In many cases, it involves the execution of resource-intensive tasks in terms of computing power such as CPU and RAM. Due to the emergence of Cloud computing, theoretically unlimited resources can be used for the enactment of business processes. These Cloud resources render several challenges for Business Process Management Systems to ensure a predefined Quality of Service level during Cloud-based process enactment. Therefore, new solutions for process scheduling and resource allocation are required to tackle these challenges. Within this paper, we present a novel approach to schedule business processes and optimize the used Cloud-based computational resources in a cost-efficient way, thus realizing so-called elastic processes. For that, we specify the Service Instance Placement Problem, i.e., an optimization model which defines the setting of how service instances are scheduled among resources. Through extensive evaluations we show the benefits of our contributions and compare the novel approach against a baseline which follows an ad hoc approach.Publicationtype: Journal Article19 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Elastic Stream Processing for Distributed EnvironmentsThe current development towards the Internet of Things introduces the need for more flexibility in stream processing. To counter these challenges, the authors propose elastic stream processing for distributed environments, building on top of cloud computing and allowing a scalable and more flexible solution compared to traditional approaches.Publicationtype: Journal Article23 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Cost-Efficient Scheduling of Elastic Processes in Hybrid CloudsCloud computing is becoming increasingly important for executing business processes. This development contributes to a novel class of Business Process Management Systems, called eBPMS, that inherit elasticity from cloud computing. The aim of eBPMS is to improve the efficiency of process enactment, in particular regarding scalability and cost-efficiency. However, there is hardly any research that investigates scheduling for eBPMS so far. Against this background, we design an elastic scheduling approach for eBPMS and a corresponding formal problem definition in order to evaluate its data transfer capabilities - this is especially important for hybrid cloud environments. Through extensive evaluations, we are able to show that our approach reduces the total cost by a considerable share.Publicationtype: Conference Paper26 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Elastic process optimization : the service instance placement problem(Technische Universität Wien, 2014-10); ; ; ; Publicationtype: Technical Report27 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Towards Process Support for Cloud Manufacturing(IEEE, 2014-09); ; ; ; Due to increasing competitive pressure, manufacturing companies need to support flexible and scalable business processes-both on the shop floor and in their enterprise software systems. Cloud manufacturing is a recent approach to realize real-world manufacturing processes by applying well-known basic concepts from the field of Cloud computing to this domain. To implement Cloud manufacturing, it is necessary to model, enact and monitor according manufacturing processes and virtualize the single process steps. So far, Business Process Management Systems do not explicitly support Cloud manufacturing. This paper analyzes requirements regarding process enactment for Cloud manufacturing and provides a concept for an according software framework.Publicationtype: Conference Paper21