Matthias Schnepf

HEPScore is a new CPU benchmark created to replace the HEPSPEC06 benchmark that is currently used by the WLCG for procurement, computing resource pledges, usage accounting and performance studies. The development of the new benchmark, based on HEP applications or workloads, has involved many contributions from software developers, data analysts, experts of the experiments, representatives of several WLCG computing centres and WLCG site managers. In this contribution, we review the selection of workloads and the validation of the new HEPScore benchmark.

Many common CPU architectures provide simultaneous multithreading (SMT). The operating system sees multiple logical CPU cores per physical CPU core and can schedule several processes to one physical CPU core. This overbooking of physical cores enables a better usage of parallel pipelines and doubled components within a CPU core. On systems with several applications running in parallel, such as batch jobs on worker nodes, the usage of SMT can increase the overall performance. In high energy physics (HEP) batch/Grid jobs are accounted for in units of single-core jobs. One single-core job is designed to utilize one logical CPU core fully. As a result, Grid sites often configure their worker nodes to provide as many single-core slots as physical or logical CPU cores. However, due to memory and disk space constraints, not all logical CPU cores can be used. Therefore, it can be useful to configure more single-core slots than physical CPU cores but fewer than logical CPU cores per worker node. We have extensively used and studied this strategy at the GridKa Tier 1 center. In this contribution, we show benchmark results for different overbooking factors of physical cores on various CPU models for different HEP workflows and Benchmarks.

To satisfy future computing demands of the Worldwide LHC Computing Grid (WLCG), opportunistic usage of third-party resources is a promising approach. While the means to make such resources compatible with WLCG requirements are largely satisfied by virtual machines and containers technologies, strategies to acquire and disband many resources from many providers are still a focus of current research. Existing meta-schedulers that manage resources in the WLCG are hitting the limits of their design when tasked to manage heterogeneous resources from many diverse resource providers. To provide opportunistic resources to the WLCG as part of a regular WLCG site, we propose a new meta-scheduling approach suitable for opportunistic, heterogeneous resource provisioning. Instead of anticipating future resource requirements, our approach observes resource usage and promotes well-used resources. Following this approach, we have developed an inherently robust meta-scheduler, COBalD, for managing diverse, heterogeneous resources given unpredictable resource requirements. This paper explains the key concepts of our approach, and discusses the benefits and limitations of our new approach to dynamic resource provisioning compared to previous approaches.

Demand for computing resources in high energy physics (HEP) shows a highly dynamic behavior, while the provided resources by the Worldwide LHC Computing Grid (WLCG) remains static. It has become evident that opportunistic resources such as High Performance Computing (HPC) centers and commercial clouds are well suited to cover peak loads. However, the utilization of these resources gives rise to new levels of complexity, e.g. resources need to be managed highly dynamically and HEP applications require a very specific software environment usually not provided at opportunistic resources. Furthermore, aspects to consider are limitations in network bandwidth causing I/O-intensive workflows to run inefficiently. The key component to dynamically run HEP applications on opportunistic resources is the utilization of modern container and virtualization technologies. Based on these technologies, the Karlsruhe Institute of Technology (KIT) has developed ROCED, a resource manager to dynamically integrate and manage a variety of opportunistic resources. In combination with ROCED, HTCondor batch system acts as a powerful single entry point to all available computing resources, leading to a seamless and transparent integration of opportunistic resources into HEP computing. KIT is currently improving the resource management and job scheduling by focusing on I/O requirements of individual workflows, available network bandwidth as well as scalability. For these reasons, we are currently developing a new resource manager, called TARDIS. In this paper, we give an overview of the utilized technologies, the dynamic management, and integration of resources as well as the status of the I/O-based resource and job scheduling.

This contribution reports on solutions, experiences and recent developments with the dynamic, on-demand provisioning of remote computing resources for analysis and simulation workflows. Local resources of a physics institute are extended by private and commercial cloud sites, ranging from the inclusion of desktop clusters over institute clusters to HPC centers.

High throughput and short turnaround cycles are core requirements for efficient processing of data-intense end-user analyses in High Energy Physics (HEP). Together with the tremendously increasing amount of data to be processed, this leads to enormous challenges for HEP storage systems, networks and the data distribution to computing resources for end-user analyses. Bringing data close to the computing resource is a very promising approach to solve throughput limitations and improve the overall performance. However, achieving data locality by placing multiple conventional caches inside a distributed computing infrastructure leads to redundant data placement and inefficient usage of the limited cache volume. The solution is a coordinated placement of critical data on computing resources, which enables matching each process of an analysis work-flow to its most suitable worker node in terms of data locality and, thus, reduces the overall processing time. This coordinated distributed caching concept was realized at KIT by developing the coordination service NaviX that connects an XRootD cache proxy infrastructure with an HTCondor batch system. We give an overview about the coordinated distributed caching concept and experiences collected on prototype system based on NaviX.

Increased operational effectiveness and the dynamic integration of only temporarily available compute resources (opportunistic resources) becomes more and more important in the next decade, due to the scarcity of resources for future high energy physics experiments as well as the desired integration of cloud and high performance computing resources. This results in a more heterogenous compute environment, which gives rise to huge challenges for the computing operation teams of the experiments. At the Karlsruhe Institute of Technology (KIT) we design solutions to tackle these challenges. In order to ensure an efficient utilization of opportunistic resources and unified access to the entire infrastructure, we developed the Transparent Adaptive Resource Dynamic Integration System (TARDIS). A scalable multi-agent resource manager providing interfaces to provision as well as dynamically and transparently integrate resources of various providers into one common overlay batch system. Operational effectiveness is guaranteed by relying on COBalD – the Opportunistic Balancing Daemon and its simple approach of taking into account the utilization and allocation of the different resource types, in order to run the individual workflows on the best-suited resource respectively. In this contribution we will present the current status of integrating various HPC centers and cloud providers into the compute infrastructure at the Karlsruhe Institute of Technology as well as our experiences gained in a production environment.

As results of the excellent LHC performance in 2016, more data than expected has been recorded leading to a higher demand for computing resources. It is already foreseeable that for the current and upcoming run periods a flat computing budget and the expected technology advance will not be sufficient to meet the future requirements. This results in a growing gap between supplied and demanded resources.

The GridKa Tier 1 data and computing center hosts a significant share of WLCG processing resources. Providing these resources to all major LHC and other VOs requires efficient, scalable and reliable cluster management. To satisfy this, GridKa has recently migrated its batch resources from CREAM-CE and PBS to ARC-CE and HTCondor. This contribution discusses the key highlights of the adoption of this middleware at the scale of a European Tier 1 center: As the largest WLCG Tier 1 using the ARC-CE plus HTCondor stack, GridKa is exemplary for migrating more than 20 000 cores over the time span of only a few weeks. Supporting multiple VOs, we have extensively studied the constraints and possibilities of scheduling jobs of vastly different requirements. We present a robust and maintainable optimization of resource utilization which still respects constraints desired by VOs. Furthermore, we explore the dynamic extension of our batch system, integrating cloud resources with a lightweight configuration mechanism.

We built a cluster of ARM-based Cubieboards2 which has a SATA interface to connect a harddrive. This cluster was set up as a storage system using Ceph and as a compute cluster for high energy physics analyses. To study the performance in these applications, we ran two benchmarks on this cluster. We also checked the energy efficiency of the cluster using the preseted benchmarks. Performance and energy efficency of our cluster were compared with a network-attached storage (NAS), and with a desktop PC.

With the increasing data volume of LHC Run2, user analyses are evolving towards increasing data throughput. This evolution translates to higher requirements for efficiency and scalability of the underlying analysis infrastructure. We approach this issue with a new middleware to optimise data access: a layer of coordinated caches transparently provides data locality for high-throughput analyses. We demonstrated the feasibility of this approach with a prototype used for analyses of the CMS working groups at KIT. In this paper, we present our experience both with the approach in general, and our prototype in specific.

The heavily increasing amount of data produced by current experiments in high energy particle physics challenge both end users and providers of computing resources. The boosted data rates and the complexity of analyses require huge datasets being processed in short turnaround cycles. Usually, data storages and computing farms are deployed by different providers, which leads to data delocalization and a strong influence of the interconnection transfer rates. The CMS collaboration at KIT has developed a prototype enabling data locality for HEP analysis processing via two concepts. A coordinated and distributed caching approach that reduce the limiting factor of data transfers by joining local high performance devices with large background storages were tested. Thereby, a throughput optimization was reached by selecting and allocating critical data within user work-flows. A highly performant setup using these caching solutions enables fast processing of throughput dependent analysis workflows.

Abstract The German CMS community (DCMS) as a whole can benefit from the various compute resources, available to its different institutes. While Grid-enabled and National Analysis Facility resources are usually shared within the community, local and recently enabled opportunistic resources like HPC centers and cloud resources are not. Furthermore, there is no shared submission infrastructure available. Via HTCondor’s [1] mechanisms to connect resource pools, several remote pools can be connected transparently to the users and therefore used more efficiently by a multitude of user groups. In addition to the statically provisioned resources, also dynamically allocated resources from external cloud providers as well as HPC centers can be integrated. However, the usage of such dynamically allocated resources gives rise to additional complexity. Constraints on access policies of the resources, as well as workflow necessities have to be taken care of. To maintain a well-defined and reliable runtime environment on each resource, virtualization and containerization technologies such as virtual machines, Docker, and Singularity, are used.

Abstract Data-intensive end-user analyses in high energy physics require high data throughput to reach short turnaround cycles. This leads to enormous challenges for storage and network infrastructure, especially when facing the tremendously increasing amount of data to be processed during High-Luminosity LHC runs. Including opportunistic resources with volatile storage systems into the traditional HEP computing facilities makes this situation more complex. Bringing data close to the computing units is a promising approach to solve throughput limitations and improve the overall performance. We focus on coordinated distributed caching by coordinating workows to the most suitable hosts in terms of cached files. This allows optimizing overall processing efficiency of data-intensive workows and efficiently use limited cache volume by reducing replication of data on distributed caches. We developed a NaviX coordination service at KIT that realizes coordinated distributed caching using XRootD cache proxy server infrastructure and HTCondor batch system. In this paper, we present the experience gained in operating coordinated distributed caches on cloud and HPC resources. Furthermore, we show benchmarks of a dedicated high throughput cluster, the Throughput-Optimized Analysis-System (TOpAS), which is based on the above-mentioned concept.

Abstract The current experiments in high energy physics (HEP) have a huge data rate. To convert the measured data, an enormous number of computing resources is needed and will further increase with upgraded and newer experiments. To fulfill the ever-growing demand the allocation of additional, potentially only temporary available non-HEP dedicated resources is important. These so-called opportunistic resources cannot only be used for analyses in general but are also well-suited to cover the typical unpredictable peak demands for computing resources. For both use cases, the temporary availability of the opportunistic resources requires a dynamic allocation, integration, and management, while their heterogeneity requires optimization to maintain high resource utilization by allocating best matching resources. To find the best matching resources which should be allocated is challenging due to the unpredictable submission behavior as well as an ever-changing mixture of workflows with different requirements. Instead of predicting the best matching resource, we base our decisions on the utilization of resources. For this reason, we are developing the resource manager TARDIS (Transparent Adaptive Resource Dynamic Integration System) which manages and dynamically requests or releases resources. The decision of how many resources TARDIS has to request is implemented in COBalD (COBald - The Opportunistic Balancing Daemon) to ensure further allocation of well-used resources while reducing the amount of insufficiently used ones. TARDIS allocates and manages resources from various resource providers such as HPC centers or commercial and public clouds while ensuring a dynamic allocation and efficient utilization of these heterogeneous opportunistic resources. Furthermore, TARDIS integrates the allocated opportunistic resources into one overlay batch system which provides a single point of entry for all users. In order to provide the dedicated HEP software environment, we use virtualization and container technologies. In this contribution, we give an overview of the dynamic integration of opportunistic resources via TARDIS/COBalD in our HEP institute as well as how user analyses benefit from these additional resources.

Computing resource needs are expected to increase drastically in the future. The HEP experiments ATLAS and CMS foresee an increase of a factor of 5-10 in the volume of recorded data in the upcoming years. The current infrastructure, namely the WLCG, is not sufficient to meet the demands in terms of computing and storage resources. The usage of non HEP specific resources is one way to reduce this shortage. However, using them comes at a cost: First, with multiple of such resources at hand, it gets more and more diffcult for the single user, as each resource normally requires its own authentication and has its own way of accessing it. Second, as they are not specifically designed for HEP workflows, they might lack dedicated software or other necessary services. Allocating the resources at the different providers can be done by COBalD/TARDIS, developed at KIT. The resource manager integrates resources on demand into one overlay batch system, providing the user with a single point of entry. The software and services, needed for the communities workflows, are transparently served through containers. With this, an HPC cluster at RWTH Aachen University is dynamically and transparently integrated into a Tier 2 WLCG resource, virtually doubling its computing capacities.

The inclusion of opportunistic resources, for example from High Performance Computing (HPC) centers or cloud providers, is an important contribution to bridging the gap between existing resources and future needs by the LHC collaborations, especially for the HL-LHC era. However, the integration of these resources poses new challenges and often needs to happen in a highly dynamic manner. To enable an effective and lightweight integration of these resources, the tools COBalD and TARDIS are developed at KIT. In this contribution we report on the infrastructure we use to dynamically offer opportunistic resources to collaborations in the World Wide LHC Computing Grid (WLCG). The core components are COBalD/TARDIS, HTCondor, CVMFS and modern virtualization technology. The challenging task of managing the opportunistic resources is performed by COBalD/TARDIS. We showcase the challenges, employed solutions and experiences gained with the provisioning of opportunistic resources from several resource providers like university clusters, HPC centers and cloud setups in a multi VO environment. This work can serve as a blueprint for approaching the provisioning of resources from other resource providers.