Apache Spark is a widely adopted framework for large-scale data processing. However, in industrial analytics environments, Spark's built-in schedulers, such as FIFO and fair scheduling, struggle to maintain both user-level fairness and low mean response time, particularly in long-running shared applications. Existing solutions typically focus on job-level fairness which unintentionally favors users who submit more jobs. Although Spark offers a built-in fair scheduler, it lacks adaptability to dynamic user workloads and may degrade overall job performance. We present the User Weighted Fair Queuing (UWFQ) scheduler, designed to minimize job response times while ensuring equitable resource distribution across users and their respective jobs. UWFQ simulates a virtual fair queuing system and schedules jobs based on their estimated finish times under a bounded fairness model. To further address task skew and reduce priority inversions, which are common in Spark workloads, we introduce runtime partitioning, a method that dynamically refines task granularity based on expected runtime. We implement UWFQ within the Spark framework and evaluate its performance using multi-user synthetic workloads and Google cluster traces. We show that UWFQ reduces the average response time of small jobs by up to 74% compared to existing built-in Spark schedulers and to state-of-the-art fair scheduling algorithms.

Traditional datacenter analysis is based on high-level, coarse-grained metrics. This obscures our vision of datacenter behavior, as we do not observe the full picture nor subtleties that might make up these high-level, coarse metrics. There is room for operational improvement based on fine-grained temporal and spatial, low-level metric data. We leverage in this work one of the (rare) public datasets providing fine-grained information on datacenter operations, with over 60 billion measurements captured in 15-second intervals. We show evidence that fine-grained information reveals new operational aspects, that the different metrics cannot be derived from one another (and thus need to be captured), and that many low-level metrics, gathered frequently are key to understanding datacenter operations. We propose a holistic analysis for datacenter operations, providing statistical characterization of node and workload aspects. Our analysis reveals both generic and machine learning-specific aspects, summarized in over 30 observations, providing deep insight into this dataset and the originating cluster. We give actionable insights, surprising findings, and exemplify how our observations support performance-engineering tasks such as workload prediction and long-term datacenter design.

Cloud datacenters are important for the digital society, serving stakeholders across industry, government, and academia. Simulation is a critical part of exploring datacenter technologies, enabling scalable experimentation with millions of jobs and hundreds of thousands of machines, and what-if analysis in a matter of minutes to hours. Although the community has already developed powerful simulators, emerging technologies and applications in modern datacenters require new approaches. Addressing this requirement, in this work we propose OpenDC, a new platform for datacenter simulation. OpenDC includes novel models for emerging cloud-datacenter technologies and applications, such as serverless computing with FaaS deployment and TensorFlow-based machine learning. Our design also focuses on convenience, with a web-based interface for interactive experimentation, support for experiment automation, a library of prefabs for constructing and sharing datacenter designs, and support for diverse input formats and output metrics. We implement, validate, and open-source OpenDC 2.0, a significant redesign and release after a multi-year research and development process. We demonstrate the benefits of OpenDC for the field through a set of representative use-cases: serverless, machine learning, procurement of HPC-as-a-Service infrastructure, educational practices, and reproducibility studies. Overall, OpenDC helps understand how datacenters work, design datacenter infrastructure, and train the next generation of experts.

Workflows are prevalent in today’s computing infrastructures as they support many domains. Different Quality of Service (QoS) requirements of both users and providers makes workflow scheduling challenging. Meeting the challenge requires an overview of state-of-art in workflow scheduling. Sifting through literature to find the state-of-art can be daunting, for both newcomers and experienced researchers. Surveys are an excellent way to address questions regarding the different techniques, policies, emerging areas, and opportunities present, yet they rarely take a systematic approach and publish their tools and data on which they are based. Moreover, the communities behind these articles are rarely studied. We attempt to address these shortcomings in this work. We introduce and open-source an instrument used to combine and store article meta-data. Using this meta-data, we characterize and taxonomize the workflow scheduling community and four areas within workflow scheduling: (1) the workflow formalism, (2) workflow allocation, (3) resource provisioning, and (4) applications and services. In each characterization, we obtain important keywords overall and per year, identify keywords growing in importance, get insight into the structure and relations within each community, and perform a systematic literature survey per part to validate and complement our taxonomies

The rapid adoption and the diversification of cloud computing technology exacerbate the importance of a sound experimental methodology for this domain. This work investigates how to measure and report performance in the cloud, and how well the cloud research community is already doing it. We propose a set of eight important methodological principles that combine best-practices from nearby fields with concepts applicable only to clouds, and with new ideas about the time-accuracy trade-off. We show how these principles are applicable using a practical use-case experiment. To this end, we analyze the ability of the newly released SPEC Cloud IaaS benchmark to follow the principles, and showcase real-world experimental studies in common cloud environments that meet the principles. Last, we report on a systematic literature review including top conferences and journals in the field, from 2012 to 2017, analyzing if the practice of reporting cloud performance measurements follows the proposed eight principles. Worryingly, this systematic survey and the subsequent two-round human reviews, reveal that few of the published studies follow the eight experimental principles. We conclude that, although these important principles are simple and basic, the cloud community is yet to adopt them broadly to deliver sound measurement of cloud environments.

Realistic, relevant, and reproducible experiments often need input traces collected from real-world environments. In this work, we focus on traces of workflows-common in datacenters, clouds, and HPC infrastructures. We show that the state-of-the-art in using workflow-traces raises important issues: (1) the use of realistic traces is infrequent and (2) the use of realistic, open-access traces even more so. Alleviating these issues, we introduce the Workflow Trace Archive (WTA), an open-access archive of workflow traces from diverse computing infrastructures and tooling to parse, validate, and analyze traces. The WTA includes > 48 million workflows captured from > 10 computing infrastructures, representing a broad diversity of trace domains and characteristics. To emphasize the importance of trace diversity, we characterize the WTA contents and analyze in simulation the impact of trace diversity on experiment results. Our results indicate significant differences in characteristics, properties, and workflow structures between workload sources, domains, and fields.

Nowadays, in order to keep track of the fast changing requirements of Internet applications, auto-scaling is used as an essential mechanism for adapting the number of provisioned resources to the resource demand. The straightforward approach is to deploy a set of common and opensource single-service auto-scalers for each service independently. However, this deployment leads to problems such as bottleneck-shifting and increased oscillations. Existing auto-scalers that scale applications consisting of multiple services are kept closed-source. To face these challenges, we first survey existing auto-scalers and highlight current challenges. Then, we introduce Chamulteon, a redesign of our previously introduced mechanism, which can scale applications consisting of multiple services in a coordinated manner. We evaluate Chamulteon against four different well-cited auto-scalers in four sets of measurement-based experiments where we use diverse environments (VM vs. Docker), real-world traces, and vary the scale of the demanded resources. Overall, Chamulteon achieves the best auto-scaling performance based on established user-oriented and endorsed elasticity metrics.

High-quality designs of distributed systems and services are essential for our digital economy and society. Threatening to slow down the stream of working designs, we identify the mounting pressure of scale and complexity of (eco-)systems, of ill-defined and wicked problems, and of unclear processes, methods, and tools. We envision design itself as a core research topic in distributed systems, to understand and improve the science and practice of distributed (eco-)system design. Toward this vision, we propose the AtLarge design framework, accompanied by a set of 8 core design principles. We also propose 10 key challenges, which we hope the community can address in the following 5 years. In our experience so far, the proposed framework and principles are practical, and lead to pragmatic and innovative designs for large-scale distributed systems.

To improve customer experience, datacenter operators offer support for simplifying application and resource management. For example, running workloads of workflows on behalf of customers is desirable, but requires increasingly more sophisticated autoscaling policies, that is, policies that dynamically provision resources for the customer. Although selecting and tuning autoscaling policies is a challenging task for datacenter operators, so far relatively few studies investigate the performance of autoscaling for workloads of workflows. Complementing previous knowledge, in this work we propose the first comprehensive performance study in the field. Using trace-based simulation, we compare state-of-the-art autoscaling policies across multiple application domains, workload arrival patterns (e.g., burstiness), and system utilization levels. We further investigate the interplay between autoscaling and regular allocation policies, and the complexity cost of autoscaling. Our quantitative study focuses not only on traditional performance metrics and on state-of-the-art elasticity metrics, but also on time-and memory-related autoscaling-complexity metrics. Our main results give strong and quantitative evidence about previously unreported operational behavior, for example, that autoscaling policies perform differently across application domains and allocation and provisioning policies should be co-designed.

Our society is digital: industry, science, governance, and individuals depend, often transparently, on the inter-operation of large numbers of distributed computer systems. Although the society takes them almost for granted, these computer ecosystems are not available for all, may not be affordable for long, and raise numerous other research challenges. Inspired by these challenges and by our experience with distributed computer systems, we envision Massivizing Computer Systems, a domain of computer science focusing on understanding, controlling, and evolving successfully such ecosystems. Beyond establishing and growing a body of knowledge about computer ecosystems and their constituent systems, the community in this domain should also aim to educate many about design and engineering for this domain, and all people about its principles. This is a call to the entire community: there is much to discover and achieve.

In the new Digital Economy, massive computer systems, often grouped in datacenters, serve as factories "producing" cloud services with massive consumption. However, to afford cloud services globally, we must address new research challenges in designing, operating, and using modern datacenters. We must also address challenges in educating and training the next generation of datacenter engineers. Addressing such challenges, in this work we present our vision on OpenDC: we envision the exploration of various datacenter concepts and technologies, using existing and new scientific methods, enabling new education practices and topics, and leading to the creation of new software and data artifacts. We present the datacenter concepts and technologies we are currently planning to explore using OpenDC. We identify the scientific methods we want to use, and explain our vision of education practices. We present the architecture and open-source program underlying the OpenDC software, and the format and open-access data we use for datacenter experiments. We conclude with an open invitation for the community to join our effort.