Business processes are at each company's core and must be adapted permanently to react to changing markets, substantial growth, or legal regulations. Especially digital ventures have the potential to evolve fast, and consequently, their business processes need to change at the same speed. Two streams of literature have looked into this. Traditional business process management sees business processes, once implemented, as relatively stable. In contrast, digital entrepreneurship literature highlights the inherent flexibility of digital ventures. Based on amultiple case study of five digital ventures, we analyze how entrepreneurs deal with this tension when business processes evolve. Building on entrepreneurial bricolage, we propose two types of resource recombination that we find, namely, usage of existing private resources and re-configuring of resources already being used within the venture. These insights contribute to extending our understanding of the evolution of business processes.

Business process management (BPM) is changing in the digital age. As a result, organizations are confronted with new logics that their business processes adhere to: processes are designed to allow for easy adaptability, infrastructure becomes progressively more flexible, and process participants make their own decisions in ambiguous situations. In this context, business process change becomes increasingly important. Digital ventures – key phenomena in the digital age – heavily rely on digital technology and, hence, have the potential to change quickly. Consequently, their business processes need to change at the same speed. While the literature on BPM proposes different types of business process change and acknowledges that digital technology can enable such developments, it remains to be explored which specific characteristics of digital technology facilitate business process change. The study investigates this by drawing on a multiple case study with seven digital ventures. It finds four patterns of business process changes in digital ventures, illustrating digital technology’s impact on business processes. The study compares the patterns with existing types of business process change from the literature and discusses differences and similarities, trying to advance the understanding of business process dynamics in the digital age.

In this final chapter of the Handbook, we introduce emerging, innovative methods in software fault localization. First, we present scientific and systematic hypothesis-testing techniques and show they may be applied in practice. Second, for fault localization in the absence of a test oracle, we present a semi-proving methodology based on metamorphic relations and symbolic evaluation. It hinges on causes and effects instead of statistical probabilities. Third, we present an approach to predict the effectiveness of fault localization tools using machine learning. Lastly, we discuss why manually produced test cases are not ideal for fault localization and explain how to mitigate the problem by using automatically generated test cases.

Enterprise Resource Planning (ERP) systems are critical to the success of enterprises, facilitating business operations through standardized digital processes. However, existing ERP systems are unsuitable for startups and small and medium-sized enterprises that grow quickly and require adaptable solutions with low barriers to entry. Drawing upon 15 explorative interviews with industry experts, we examine the challenges of current ERP systems using the task technology fit theory across companies of varying sizes. We describe high entry barriers, high costs of implementing implicit processes, and insufficient interoperability of already employed tools. We present a vision of a future business process platform based on three enablers: Business processes as first-class entities, semantic data and processes, and cloud-native elasticity and high availability. We discuss how these enablers address current ERP systems' challenges and how they may be used for research on the next generation of business software for tomorrow's enterprises.

To combat the trend of failing Digital Transformation endeavors, dedicated leadership skills are needed. However, so far there is little knowledge about the skills for successfully leading Digital Innovation Units (DIUs). We, therefore, interviewed 13 DIU leaders from various industries and elicited their skills with the Repertory Grid method. We identified 54 key skills clustered in seven categories: team development, integration of the DIU into the wider context of the organization, innovation management, personal traits, effective communication, hard skills, and visionary thinking and driving change. Furthermore, we found five influencing factors for the application of key skills: Financial support, uncertainty, certainty in tasks, stability of the product, freedom to operate, and degree of collaboration with the core organization. Finally, we report on three skill enhancers (actions, attitudes, conditions) for practitioners. We complement prior research by improving the understanding of the ideal DIU leader skill profile.

The Hasso Plattner Institute (HPI), academically structured as the independent Faculty of Digital Engineering at the University of Potsdam, unites computer science research and teaching with the advantages of a privately financed institute and a tuition-free study program. Founder and namesake of the institute is the SAP co-founder Hasso Plattner, who also heads the Enterprise Platform and Integration Concepts (EPIC) research center which focuses on the technical aspects of business software with a vision to provide the fastest way to get insights out of enterprise data. Founded in 2006, the EPIC combines three research groups comprising autonomous data management, enterprise software engineering, and data-driven decision support.

Memory allocation has a significant impact on the performance of in-memory databases. While state-of-the-art memory allocators work well in DRAM-only setups, some of their design decisions might no longer yield efficiency if data is tiered to disaggregated memory or secondary memory tiers. In this work, we study the performance impact of metadata in memory allocators and their tiering to disaggregated memory in the context of in-memory databases for the first time. We show how to separate metadata and application data by the example of jemalloc, which is widely used for data-intensive applications, and study performance effects for different workloads.

COVID-19 and its immediate impacts on teaching activities have required changes from computer science educators worldwide.We switched our on-site courses to remote setups without detailed knowledge of what tools, techniques, and methods would work in different teaching contexts.A growing amount of experience reports on general best practices for remote teaching in higher education are available.However, university courses featuring practical software development projects present unique challenges regarding remote learning, as effective student collaboration is vital.In these courses, students tackle situations in the project and their team meetings that would also occur in real software projects experienced in industry settings.In this paper, we share our experiences on how we successfully adapted our software engineering project course to a remote setup, which challenges we observed in student teams and how they can be mitigated, and what (surprisingly) worked better than expected.Finally, we propose improvements that we expect will be beneficial not only for future remote-only but also for hybrid or on-site courses.

Das 1998 gegründete Hasso-Plattner-Institut (HPI) ist ein privat finanziertes IT-Institut und bildet gemeinsam mit der Universität Potsdam die Fakultät „Digital Engineering“. Gründer und Namensgeber des Instituts ist der SAP-Mitgründer Hasso Plattner, welcher mit dem Fachgebiet „Enterprise Platform und Integration Concepts“ (EPIC) seine eigene Forschungsgruppe leitet. Diese beschäftigt sich mit der effizienten Verwaltung, Integration und Auswertung von Unternehmens- und Geschäftsprozessdaten. Dabei wird eng mit verschiedenen Unternehmen und deren Nutzer:innen zusammengearbeitet, um Herausforderungen zu identifizieren und Lösungen gemeinsam zu entwickeln. Gegründet wurde das EPIC-Fachgebiet 2006 und bietet zurzeit zwei Senior-Researchern, drei Postdoktoranden, 16 Doktorand:innen und 21 Masterstudent:innen eine Forschungsumgebung.

The examination of causal structures is crucial for data scientists in a variety of machine learning application scenarios. In recent years, the corresponding interest in methods of causal structure learning has led to a wide spectrum of independent implementations, each having specific accuracy characteristics and introducing implementation-specific overhead in the runtime. Hence, considering a selection of algorithms or different implementations in different programming languages utilizing different hardware setups becomes a tedious manual task with high setup costs. Consequently, a tool that enables to plug in existing methods from different libraries into a single system to compare and evaluate the results is substantial support for data scientists in their research efforts. In this work, we propose an architectural blueprint of a pipeline for causal structure learning and outline our reference implementation MPCSL that addresses the requirements towards platform independence and modularity while ensuring the comparability and reproducibility of experiments. Moreover, we demonstrate the capabilities of MPCSL within a case study, where we evaluate existing implementations of the well-known PC-Algorithm concerning their runtime performance characteristics.

Databases are highly optimized complex systems with a multitude of configuration options. Especially in cloud scenarios with thousands of database deployments, determining optimized database configurations in an automated fashion is of increasing importance for database providers. At the same time, due to increased system complexity, it becomes more challenging to identify well-performing configurations. Therefore, research interest in autonomous or self-driving database systems has increased enormously in recent years. Such systems promise both performance improvements and cost reductions. In the literature, various fully or partially autonomous optimization mechanisms exist that optimize single aspects, e.g., index selection. However, database administrators and developers often distrust autonomous approaches, and there is a lack of practical experimentation opportunities that could create a better understanding. Moreover, the interplay of different autonomous mechanisms under complex workloads remains an open question. The presented cockpit enables an interactive assessment of the impact of autonomous components for database systems by comparing (autonomous) systems with different configurations side by side. Thereby, the cockpit enables users to build trust in autonomous solutions by experimenting with such technologies and observing their effects in practice.

Growing data volumes and velocities in fields such as Industry 4.0 or the Internet of Things have led to the increased popularity of data stream processing systems. Enterprises can leverage these developments by enriching their core business data and analyses with up-to-date streaming data. Comparing streaming architectures for these complex use cases is challenging, as existing benchmarks do not cover them. ESPBench is a new enterprise stream processing benchmark that fills this gap. We present its architecture, the benchmarking process, and the query workload. We employ ESPBench on three state-of-the-art stream processing systems, Apache Spark, Apache Flink, and Hazelcast Jet, using provided query implementations developed with Apache Beam. Our results highlight the need for the provided ESPBench toolkit that supports benchmark execution, as it enables query result validation and objective latency measures.

The most prominent Agile framework Scrum, is often criticized for its amount of meetings. These regular events are essential to the empirical inspect-and-adapt cycle proposed by Agile methods. Scrum meetings face several challenges, such as being perceived as boring, repetitive, or irrelevant, leading to decreased cooperation in teams and less successful projects. In an attempt to address these challenges, Agile practitioners have adopted teamwork, innovation, and design techniques geared towards improving collaboration. Additionally, they have developed their own activities to be used in Scrum meetings, most notably for conducting retrospective and planning events. Design thinking incorporates non-designers and designers in design and conceptualization activities, including user research, ideation, or testing. Accordingly, the design thinking approach provides a process with different phases and accompanying techniques for each step. These design thinking techniques can support shared understanding in teams and can improve collaboration, creativity, and product understanding. For these reasons, design thinking techniques represent a worthwhile addition to the Scrum meeting toolkit and can support Agile meetings in preventing or countering common meeting challenges and achieving meeting goals. This chapter explores how techniques from the design thinking toolkit can support Scrum meetings from a theoretical and practical viewpoint. We analyze Scrum meetings' requirements, goals, and challenges and link them to groups of techniques from the design thinking toolkit. In addition, we review interview and observational data from two previous studies with software development practitioners and derive concrete examples. As a result, we present initial guidelines on integrating design thinking techniques into Scrum meetings to make them more engaging, collaborative, and interactive.

Lightweight data compression algorithms are often used to decrease memory consumption of in-memory databases. In recent years, various integer compression techniques have been proposed that focus on sequential encoding and decoding and exploit modern CPUs’ vectorization capabilities. Interestingly, another dominant access pattern in databases systems has seen little attention: random access decoding. In this paper, we compare end-to-end database performance for various integer compression codecs on three recent CPU architectures. Our evaluation suggests that random access performance is often more relevant than vectorization capabilities for sequential accesses. Before integrating selected encodings in the database core, we benchmarked seven libraries in an exhaustive standalone comparison. We integrated the most promising techniques into the relational in-memory database system Hyrise and evaluated their performance for TPC-H, TPC-DS, and the Join Order Benchmark on three different CPU architectures. Our results emphasize the importance of random access decoding. Compared to state-of-the-art dictionary encoding in TPC-H, alternatives allow reducing memory consumption of integer columns by up to 53 % while improving runtime performance by 5 % on an Intel CPU and over 16 % on an Apple M1. 1 COMPRESSION IN IN-MEMORY DATABASE SYSTEMS With increasing volumes of data being collected, the need for fast and efficient processing of workloads increases continuously. When customers’ workloads have substantial performance requirements that the Database Management System (DBMS) needs to serve, storing data on disk is often too slow. Technological advances in the main memory industry have made large main memory capacities affordable, enabling the adoption of in-memory databases [17]. For these DBMS, disk access is no This work is licensed under the Creative Commons BY-NC-ND 4.0 International License and appears in ADMS 2021, 12th International Workshop on Accelerating Analytics and Data Management Systems, August 16, 2021, Copenhagen, Denmark. longer the bottleneck. Instead, main memory access and processing efficiency are the new optimization goals [29]. Therefore, customers often choose to use in-memory databases to serve their performance-critical workloads. However, for in-memory databases, a compact representation of data is even more crucial. In a recent survey by the market researcher IDG [19], 80 % of the participating organizations stated that they have at least one part of their infrastructure running in the cloud and spend around one-third of their IT budget on cloud computing. Here, the bill depends on the used resources, which is often measured in RAM size1. To adapt in-memory databases to run efficiently on cloud infrastructures, therefore, means that they should use as few main memory as possible. Data compression can help to achieve this. However, several practical considerations limit the usage of compression. In addition to the strong performance requirements, Service Level Agreements (SLAs) in enterprise contexts often bound the maximum allowed query response time [30], which makes the usage of heavyweight compression schemes (e.g., LZ4 [9], LZ77 [38]) difficult. As an alternative, lightweight encoding schemes can still 75

With back-in-time debuggers, developers can explore what happened before observable failures by following infection chains back to their root causes. While there are several such debuggers for object-oriented programming languages, we do not know of any back-in-time capabilities at the database-level. Thus, if failures are caused by SQL scripts or stored procedures, developers have difficulties in understanding their unexpected behavior. In this paper, we present an approach for bringing back-in-time debugging down to the SAP HANA in-memory database. Our TARDISP debugger allows developers to step queries backwards and inspecting the database at previous and arbitrary points in time. With the help of a SQL extension, we can express queries covering a period of execution time within a debugging session and handle large amounts of data with low overhead on performance and memory. The entire approach has been evaluated within a development project at SAP and shows promising results with respect to the gathered developer feedback.

Distributed applications are hard to debug because timing-dependent network communication is a source of non-deterministic behavior. Current approaches to debug non-deterministic failures include post-mortem debugging as well as record and replay. However, the first impairs system performance to gather data, whereas the latter requires developers to understand the timing-dependent communication at a lower level of abstraction than they develop at. Furthermore, both approaches require intrusive core library modifications to gather data from live systems.In this paper, we present the Peek-At-Talk debugger for investigating non-deterministic failures with low overhead in a systematic, top-down method, with a particular focus on tool-building issues in the following areas: First, we show how our debugging framework Path Tools guides developers from failures to their root causes and gathers run-time data with low overhead. Second, we present Peek-At-Talk, an extension to our Path Tools framework to record non-deterministic communication and refine behavioral data that connects source code with network events. Finally, we scope changes to the core library to record network communication without impacting other network applications.

In 1997, Henry Lieberman stated that debugging is the dirty little secret of computer science. Since then, several promising debugging technologies have been developed such as back-in-time debuggers and automatic fault localization methods. However, the last study about the state-of-the-art in debugging is still more than 15 years old and so it is not clear whether these new approaches have been applied in practice or not. For that reason, we investigate the current state of debugging in a new comprehensive study. First, we review the available literature and learn about current approaches and study results. Second, we observe several professional developers while debugging and interview them about their experiences. Based on these results, we create a questionnaire that should serve as the basis for a large-scale online debugging survey later on. With these results, we expect new insights into debugging practice that help to suggest new directions for future research.