A5034546612- Software System Performance and Reliability
- Cloud Computing and Resource Management
- Advanced Software Engineering Methodologies
- Service-Oriented Architecture and Web Services
- Software Reliability and Analysis Research
- Distributed systems and fault tolerance
- IoT and Edge/Fog Computing
- Software Engineering Research
- Traffic Prediction and Management Techniques
- Data Stream Mining Techniques
- Distributed and Parallel Computing Systems
- Software-Defined Networks and 5G
- Software Engineering Techniques and Practices
- Customer churn and segmentation
- Business Process Modeling and Analysis
- Peer-to-Peer Network Technologies
- Advanced Queuing Theory Analysis
- Simulation Techniques and Applications
- Context-Aware Activity Recognition Systems
- Mobile Agent-Based Network Management
Regierungspräsidium Stuttgart
2022
IBM (Germany)
2020-2021
University of Würzburg
2014-2018
Karlsruhe Institute of Technology
2010-2014
Center for Information Technology
2012
Auto-scalers for clouds promise stable service quality at low costs when facing changing workload intensity. The major public cloud providers provide trigger-based auto-scalers based on thresholds. However, auto-scaling has reaction times in the order of minutes. Novel from literature try to overcome limitations reactive mechanisms by employing proactive prediction methods. adoption production is still very due high risk relying a single method. This paper tackles challenge reducing this...
DevOps is a trend towards tighter integration between development (Dev) and operations (Ops) teams. The need for such an driven by the requirement to continuously adapt enterprise applications (EAs) changes in business environment. As of today, concepts have been primarily introduced ensure constant flow features bug fixes into new releases from functional perspective. In order integrate non-functional perspective these this report focuses on tools, activities, processes one most important...
Modern IT systems have increasingly distributed and dynamic architectures providing flexibility to adapt changes in the environment thus enabling higher resource efficiency. However, these benefits come at cost of system complexity dynamics. Thus, engineering that manage their end-to-end application performance efficiency an autonomic manner is a challenge. In this article, we present holistic model-based approach for self-aware management leveraging Descartes Modeling Language (DML),...
Applications in virtualized data centers are often subject to Service Level Objectives (SLOs) regarding their performance (e.g., latency or throughput). In order fulfill these SLOs, it is necessary allocate sufficient resources of different types (CPU, memory, I/O, etc.) an application. However, the relationship between application and resource allocation complex depends on multiple factors including architecture, system configuration, workload demands. this paper, we present a model-based...
Multi-tenancy is an approach to share one application instance among multiple customers by providing each of them a dedicated view. This commonly used SaaS providers reduce the costs for service provisioning. Tenants also expect be isolated in terms performance they observe and inability offer guarantees major obstacle potential cloud customers. To guarantee it essential control resources tenant. challenge, because layers execution environment, responsible controlling resource usage(e.g.,...
When creating a performance model, it is necessary to quantify the amount of resources consumed by an application serving individual requests. In distributed enterprise systems, these resource demands usually cannot be observed directly, their estimation major challenge. Different statistical approaches demand based on monitoring data have been proposed, e.g., using linear regression or Kalman filtering techniques. this paper, we present LibReDE, library ready-to-use implementations that can...
Performance models can support decisions throughout the life-cycle of a software system. However, manual construction such performance is complex and time-consuming task requiring deep system knowledge. Therefore, automatic approaches for creating updating running are necessary. Existing work focuses on single aspects model extraction or proposes specifically designed certain technology stack. In virtualized environments, we often see different applications based diverse stacks sharing same...
Enterprise applications in virtualized environments are often subject to time-varying workloads with multiple seasonal patterns and trends. In order ensure quality of service for such while avoiding over-provisioning, resources need be dynamically adapted accommodate the current workload demands. Many memory-intensive not suitable traditional horizontal scaling approach used runtime performance management, as it relies on complex expensive state replication. On other hand, vertical memory...
Queueing Petri nets are a powerful formalism that can be exploited for modeling distributed systems and evaluating their performance scalability. By combining the power expressiveness of queueing networks stochastic nets, provide number advantages. This tutorial presents an introduction to first introducing itself then summarizing results several case studies which demonstrate how used analysis. As part tutorial, we present QPME (Queueing net Modeling Environment), open-source tool analysis...
The average time a resource needs to process incoming requests in monitored workload mix is key parameter of stochastic performance models. Direct measurement these demands usually infeasible due instrumentation overheads causing interferences and perturbation production environments.Thus, number statistical estimation approaches (e.g., based on optimization, regression or Kalman filters) have been proposed the literature each coming with different strengths run-time overheads. Most offer...
Resource demands are crucial parameters for modeling and predicting the performance of software systems. Currently, resource demand estimators usually executed once system analysis. However, monitored system, as well itself, subject to constant change in runtime environments. These changes additionally impact applicability, required parametrization resulting accuracy individual estimation approaches. Over time, this leads invalid or outdated estimates, which turn negatively influence...
Queueing Petri Nets (QPNs) are a powerful formalism to model the performance of software systems. Such models can be solved using analytical or simulation techniques. Analytical techniques suffer from scalability issues, whereas often require very long runs. Exis
Due to the growing size of modern IT systems, their performance analysis becomes an even more challenging task. Existing simulators are unable analyze behavior large systems in a reasonable time, whereas analytical methods suffer from state space explosion problem. Fluid techniques can be used approximate solution high-order Markov chain models enabling time efficient models. In this paper, we describe model-to-model transformation queueing Petri nets (QPN) into layered networks (LQN)....