Tuning configurations of Spark jobs is not a trivial task. State-of-the-art auto-tuning systems are based on iteratively running workloads with different configurations. During the optimization process, the relevant features are explored to find good solutions. Many optimizers enhance the time-to-solution using black-box optimization algorithms that do not take into account any information from the Spark workloads. In this article, we present a new method for tuning configurations that uses information from one run of a Spark workload. To achieve good performance, we mine the SparkEventLog that is generated by the Spark engine. This log file contains a large amount of information from the executed application. We use this information to enhance a performance model with low-level features from the workload to be optimized. These features include Spark Actions, Transformations, and Task metrics. This process allows us to obtain application-specific workload information. With this information our system can predict sensible Spark configurations for unseen jobs, given that it has been trained with reasonable coverage of Spark applications. Experiments show that the presented system correctly produces good configurations, while achieving up to 80% speedup with respect to the default Spark configuration, and up to 12x speedup of the time-to-solution with respect to a standard Bayesian Optimization procedure.

Productivity from day one on supercomputers that leverage new technologies requires significant preparation. An institution that procures a novel system architecture often lacks sufficient institutional knowledge and skills to prepare for it. Thus, the "Center of Excellence" (CoE) concept has emerged to prepare for systems such as Summit and Sierra, currently the top two systems in the Top 500. This paper documents CoE experiences that prepared a workload of diverse applications and math libraries for a heterogeneous system. We describe our approach to this preparation, including our management and execution strategies, and detail our experiences with and reasons for using different programming approaches. Our early science and performance results show that the project enabled significant early seismic science with up to a l4X throughput increase over Cori. In addition to our successes, we discuss our challenges and failures so others may benefit from our experience.

Computational models can define the functional dynamics of complex systems in exceptional detail. However, many modeling studies face seemingly incommensurate requirements: to gain meaningful insights into some phenomena requires models with high resolution (microscopic) detail that must nevertheless evolve over large (macroscopic) length- and time-scales. Multiscale modeling has become increasingly important to bridge this gap. Executing complex multiscale models on current petascale computers with high levels of parallelism and heterogeneous architectures is challenging. Many distinct types of resources need to be simultaneously managed, such as GPUs and CPUs, memory size and latencies, communication bottlenecks, and filesystem bandwidth. In addition, robustness to failure of compute nodes, network, and filesystems is critical.

Big data analytics is an indispensable tool in transforming science, engineering, medicine, health-care, finance and ultimately business itself. With the explosion of data sizes and need for shorter time-to-solution, in-memory platforms such as Apache Spark gain increasing popularity. In this context, data shuffling, a particularly difficult transformation pattern, introduces important challenges. Specifically, data shuffling is a key component of complex computations that has a major impact on the overall performance and scalability. Thus, speeding up data shuffling is a critical goal. To this end, state-of-the-art solutions often rely on overlapping the data transfers with the shuffling phase. However, they employ simple mechanisms to decide how much data and where to fetch it from, which leads to sub-optimal performance and excessive auxiliary memory utilization for the purpose of prefetching. The latter aspect is a growing concern, given evidence that memory per computation unit is continuously decreasing while interconnect bandwidth is increasing. This paper contributes a novel shuffle data transfer strategy that addresses the two aforementioned dimensions by dynamically adapting the prefetching to the computation. We implemented this novel strategy in Spark, a popular inmemory data analytics framework. To demonstrate the benefits of our proposal, we run extensive experiments on an HPC cluster with large core count per node. Compared with the default Spark shuffle strategy, our proposal shows: up to 40 percent better performance with 50 percent less memory utilization for buffering and excellent weak scalability.

Big data analytics is an indispensable tool in transforming science, engineering, medicine, healthcare, finance and ultimately business itself. With the explosion of data sizes and need for shorter time-to-solution, in-memory platforms such as Apache Spark gain increasing popularity. However, this introduces important challenges, among which data shuffling is particularly difficult: on one hand it is a key part of the computation that has a major impact on the overall performance and scalability so its efficiency is paramount, while on the other hand it needs to operate with scarce memory in order to leave as much memory available for data caching. In this context, efficient scheduling of data transfers such that it addresses both dimensions of the problem simultaneously is non-trivial. State-of-the-art solutions often rely on simple approaches that yield sub optimal performance and resource usage. This paper contributes a novel shuffle data transfer strategy that dynamically adapts to the computation with minimal memory utilization, which we briefly underline as a series of design principles.

Today's HPC systems use two mechanisms to address main-memory errors. Error-correcting codes make correctable errors transparent to software, while checkpoint/restart (CR) enables recovery from uncorrectable errors. Unfortunately, CR overhead will be enormous at exascale due to the high failure rate of memory. We propose a new OS-based approach that proactively avoids memory errors using prediction. This scheme exposes correctable error information to the OS, which migrates pages and off lines unhealthy memory to avoid application crashes. We analyze memory error patterns in extensive logs from a BG/P system and show how correctable error patterns can be used to identify memory likely to fail. We implement a proactive memory management system on BG/Q by extending the firmware and Linux. We evaluate our approach with a realistic workload and compare our overhead against CR. We show improved resilience with negligible performance overhead for applications.