Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 68
Encontrada(s) 1 página(s)

UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Database integrated analytics using R : initial experiences with SQL-Server + R

  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Poggi, Nicolas
© 2016 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Most data scientists use nowadays functional or semi-functional languages like SQL, Scala or R to treat data, obtained directly from databases. Such process requires to fetch data, process it, then store again, and such process tends to be done outside the DB, in often complex data-flows. Recently, database service providers have decided to integrate “R-as-a-Service” in their DB solutions. The analytics engine is called directly from the SQL query tree, and results are returned as part of the same query. Here we show a first taste of such technology by testing the portability of our ALOJA-ML analytics framework, coded in R, to Microsoft SQL-Server 2016, one of the SQL+R solutions released recently. In this work we discuss some data-flow schemes for porting a local DB + analytics engine architecture towards Big Data, focusing specially on the new DB Integrated Analytics approach, and commenting the first experiences in usability and performance obtained from such new services and capabilities., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

A Quick View on Current Techniques and Machine Learning Algorithms for Big Data Analytics

  • Berral García, Josep Lluís|||0000-0003-3037-3580
Big-data is an excellent source of knowledge and information from our systems and clients, but dealing with such amount of data requires automation, and this brings us to data mining and machine leaming techniques. In the ICT sector, as in many other sectors of research and industry, platforms and tools are being served and developed in order to help professionals to treat their data and leam from it automatically ; most of those platforms coming from big companies like Google or Microsoft , or from incubators at the Apache Foundation. This brief review explains the basics of machine learning with sorne ICT examples, and enurnerates sorne (but not all) of the most used tools for analyzing and modelling big-data., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

DRMaestro: orchestrating disaggregated resources on virtualized data-centers

  • Carneiro do Amaral, Marcelo
  • Polo Bardés, Jorda
  • Carrera Pérez, David|||0000-0003-4898-3424
  • González, Nelson
  • Yang, Chih-Chieh
  • Morari, Alessandro
  • D'Amora, Bruce
  • Youssef, Alaa
  • Steinder, Malgorzata
Modern applications demand resources at an unprecedented level. In this sense, data-centers are required to scale efficiently to cope with such demand. Resource disaggregation has the potential to improve resource-efficiency by allowing the deployment of workloads in more flexible ways. Therefore, the industry is shifting towards disaggregated architectures, which enables new ways to structure hardware resources in data centers. However, determining the best performing resource provisioning is a complicated task. The optimality of resource allocation in a disaggregated data center depends on its topology and the workload collocation. This paper presents DRMaestro, a framework to orchestrate disaggregated resources transparently from the applications. DRMaestro uses a novel flow-network model to determine the optimal placement in multiple phases while employing best-efforts on preventing workload performance interference. We first evaluate the impact of disaggregation regarding the additional network requirements under higher network load. The results show that for some applications the impact is minimal, but other ones can suffer up to 80% slowdown in the data transfer part. After that, we evaluate DRMaestro via a real prototype on Kubernetes and a trace-driven simulation. The results show that DRMaestro can reduce the total job makespan with a speedup of up to ≈1.20x and decrease the QoS violation up to ≈2.64x comparing with another orchestrator that does not support resource disaggregation., This project is supported by the IBM/BSC Technology Center for Supercomputing collaboration agreement. It has also received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

A resilient and distributed near real-time traffic forecasting application for Fog computing environments

  • Pérez, Juan Luis
  • Gutiérrez Torre, Alberto|||0000-0002-5548-3359
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Carrera Pérez, David|||0000-0003-4898-3424
In this paper we propose an architecture for a city-wide traffic modeling and prediction service based on the Fog Computing paradigm. The work assumes an scenario in which a number of distributed antennas receive data generated by vehicles across the city. In the Fog nodes data is collected, processed in local and intermediate nodes, and finally forwarded to a central Cloud location for further analysis. We propose a combination of a data distribution algorithm, resilient to back-haul connectivity issues, and a traffic modeling approach based on deep learning techniques to provide distributed traffic forecasting capabilities. In our experiments, we leverage real traffic logs from one week of Floating Car Data (FCD) generated in the city of Barcelona by a road-assistance service fleet comprising thousands of vehicles. FCD was processed across several simulated conditions, ranging from scenarios in which no connectivity failures occurred in the Fog nodes, to situations with long and frequent connectivity outage periods. For each scenario, the resilience and accuracy of both the data distribution algorithm, and the learning methods were analyzed. Results show that the data distribution process running in the Fog nodes is resilient to back-haul connectivity issues and is able to deliver data to the Cloud location even in presence of severe connectivity problems. Additionally, the proposed traffic modeling and forecasting method exhibits better behavior when run distributed in the Fog instead of centralized in the Cloud, especially when connectivity issues occur that force data to be delivered out of order to the Cloud., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

A resilient and distributed near real-time traffic forecasting application for Fog computing environments

  • Pérez, Juan L.
  • Gutierrez-Torre, Alberto
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Carrera, David
In this paper we propose an architecture for a city-wide traffic modeling and prediction service based on the Fog Computing paradigm. The work assumes an scenario in which a number of distributed antennas receive data generated by vehicles across the city. In the Fog nodes data is collected, processed in local and intermediate nodes, and finally forwarded to a central Cloud location for further analysis. We propose a combination of a data distribution algorithm, resilient to back-haul connectivity issues, and a traffic modeling approach based on deep learning techniques to provide distributed traffic forecasting capabilities. In our experiments, we leverage real traffic logs from one week of Floating Car Data (FCD) generated in the city of Barcelona by a road-assistance service fleet comprising thousands of vehicles. FCD was processed across several simulated conditions, ranging from scenarios in which no connectivity failures occurred in the Fog nodes, to situations with long and frequent connectivity outage periods. For each scenario, the resilience and accuracy of both the data distribution algorithm, and the learning methods were analyzed. Results show that the data distribution process running in the Fog nodes is resilient to back-haul connectivity issues and is able to deliver data to the Cloud location even in presence of severe connectivity problems. Additionally, the proposed traffic modeling and forecasting method exhibits better behavior when run distributed in the Fog instead of centralized in the Cloud, especially when connectivity issues occur that force data to be delivered out of order to the Cloud., This project is partially supported by the European Research Council (ERC), Spain under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493). The authors gratefully acknowledge the Reial Automvil Club de Catalunya (RACC) for the dataset of Floating Car Data provided., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Automatic generation of workload profiles using unsupervised learning pipelines

  • Buchaca Prats, David
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Carrera Pérez, David|||0000-0003-4898-3424
The complexity of resource usage and power consumption on cloud-based applications makes the understanding of application behavior through expert examination difficult. The difficulty increases when applications are seen as “black boxes”, where only external monitoring can be retrieved. Furthermore, given the different amount of scenarios and applications, automation is required. Here we examine and model application behavior by finding behavior phases. We use Conditional Restricted Boltzmann Machines (CRBM) to model time-series containing resources traces measurements like CPU, Memory and IO. CRBMs can be used to map a given given historic window of trace behaviour into a single vector. This low dimensional and time-aware vector can be passed through clustering methods, from simplistic ones like k-means to more complex ones like those based on Hidden Markov Models (HMM). We use these methods to find phases of similar behaviour in the workloads. Our experimental evaluation shows that the proposed method is able to identify different phases of resource consumption across different workloads. We show that the distinct phases contain specific resource patterns that distinguish them., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Considerations in using OpenCL on GPUs and FPGAs for throughput-oriented genomics workloads

  • Cadenelli, Nicola
  • Jaksic, Zoran
  • Polo Bardés, Jordà
  • Carrera Pérez, David|||0000-0003-4898-3424
The recent upsurge in the available amount of health data and the advances in next-generation sequencing are setting the ground for the long-awaited precision medicine. To process this deluge of data, bioinformatics workloads are becoming more complex and more computationally demanding. For this reasons they have been extended to support different computing architectures, such as GPUs and FPGAs, to leverage the form of parallelism typical of each of such architectures. The paper describes how a genomic workload such as k-mer frequency counting that takes advantage of a GPU can be offloaded to one or even more FPGAs. Moreover, it performs a comprehensive analysis of the FPGA acceleration comparing its performance to a non-accelerated configuration and when using a GPU. Lastly, the paper focuses on how, when using accelerators with a throughput-oriented workload, one should also take into consideration both kernel execution time and how well each accelerator board overlaps kernels and PCIe transferred. Results show that acceleration with two FPGAs can improve both time- and energy-to-solution for the entire accelerated part by a factor of 1.32x. Per contra, acceleration with one GPU delivers an improvement of 1.77x in time-to-solution but of a lower 1.49x in energy-to-solution due to persistently higher power consumption. The paper also evaluates how future FPGA boards with components (i.e., off-chip memory and PCIe) on par with those of the GPU board could provide an energy-efficient alternative to GPUs., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Burst-aware predictive autoscaling for containerized microservices

  • Abdullah, Muhammad
  • Iqbal, Waheed
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Polo Bardés, Jorda
  • Carrera Pérez, David|||0000-0003-4898-3424
Autoscaling methods are used for cloud-hosted applications to dynamically scale the allocated resources for guaranteeing Quality-of-Service (QoS). The public-facing application serves dynamic workloads, which contain bursts and pose challenges for autoscaling methods to ensure application performance. Existing State-of-the-art autoscaling methods are burst-oblivious to determine and provision the appropriate resources. For dynamic workloads, it is hard to detect and handle bursts online for maintaining application performance. In this article, we propose a novel burst-aware autoscaling method which detects burst in dynamic workloads using workload forecasting, resource prediction, and scaling decision making while minimizing response time service-level objectives (SLO) violations. We evaluated our approach through a trace-driven simulation, using multiple synthetic and realistic bursty workloads for containerized microservices, improving performance when comparing against existing state-of-the-art autoscaling methods. Such experiments show an increase of × 1.09 in total processed requests, a reduction of × 5.17 for SLO violations, and an increase of × 0.767 cost as compared to the baseline method., This work was partially supported by the European Research Council (ERC) under the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitiveness (TIN2015-65316-P and IJCI2016-27485) and the Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Accelerating K-mer Frequency Counting with GPU and Non-Volatile Memory

  • Cadenelli, Nicola
  • Polo Bardés, Jordà
  • Carrera, David
The emergence of Next Generation Sequencing (NGS) platforms has increased the throughput of genomic sequencing and in turn the amount of data that needs to be processed, requiring highly efficient computation for its analysis. In this context, modern architectures including accelerators and non-volatile memory are essential to enable the mass exploitation of these bioinformatics workloads. This paper presents a redesign of the main component of a state-of-the-art reference-free method for variant calling, SMUFIN, which has been adapted to make the most of GPUs and NVM devices. SMUFIN relies on counting the frequency of k-mers (substrings of length k) in DNA sequences, which also constitutes a well-known problem for many bioinformatics workloads, such as genome assembly. We propose techniques to improve the efficiency of k-mer counting and to scale-up workloads like SMUFIN that used to require 16 nodes of Marenostrum 3 to a single machine with a GPU and NVM drives. Results show that although the single machine is not able to improve the time to solution of 16 nodes, its CPU time is 7.5x shorter than the aggregate CPU time of the 16 nodes, with a reduction in energy consumption of 5.5x., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493). We are also grateful to SandDisk for lending the FusionIO cards and to Nvidia who donated the Tesla K40c., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Adaptive sliding windows for improved estimation of data center resource utilization

  • Baig, Shuja-ur-Rehman
  • Iqbal, Waheed
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Carrera Pérez, David|||0000-0003-4898-3424
Accurate prediction of data center resource utilization is required for capacity planning, job scheduling, energy saving, workload placement, and load balancing to utilize the resources efficiently. However, accurately predicting those resources is challenging due to dynamic workloads, heterogeneous infrastructures, and multi-tenant co-hosted applications. Existing prediction methods use fixed size observation windows which cannot produce accurate results because of not being adaptively adjusted to capture local trends in the most recent data. Therefore, those methods train on large fixed sliding windows using an irrelevant large number of observations yielding to inaccurate estimations or fall for inaccuracy due to degradation of estimations with short windows on quick changing trends. In this paper we propose a deep learning-based adaptive window size selection method, dynamically limiting the sliding window size to capture the trend for the latest resource utilization, then build an estimation model for each trend period. We evaluate the proposed method against multiple baseline and state-of-the-art methods, using real data-center workload data sets. The experimental evaluation shows that the proposed solution outperforms those state-of-the-art approaches and yields 16 to 54% improved prediction accuracy compared to the baseline methods., This work is partially supported by the European ResearchCouncil (ERC) under the EU Horizon 2020 programme(GA 639595), the Spanish Ministry of Economy, Industry andCompetitiveness (TIN2015-65316-P and IJCI2016-27485), theGeneralitat de Catalunya, Spain (2014-SGR-1051) and Universityof the Punjab, Pakistan. The statements made herein are solelythe responsibility of the authors., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Real-time data center's telemetry reduction and reconstruction using Markov chain models

  • Baig, Shuja-ur-Rehman
  • Iqbal, Waheed
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Erradi, Abdelkarim
  • Carrera Pérez, David|||0000-0003-4898-3424
Large-scale data centers are composed of thousands of servers organized in interconnected racks to offer services to users. These data centers continuously generate large amounts of telemetry data streams (e.g., hardware utilization metrics) used for multiple purposes, including resource management, workload characterization, resource utilization prediction, capacity planning, and real-time analytics. These telemetry streams require costly bandwidth utilization and storage space, particularly at medium-long term for large data centers. This paper addresses this problem by proposing and evaluating a system to efficiently reduce bandwidth and storage for telemetry data through real-time modeling using Markov chain based methods. Our proposed solution was evaluated using real telemetry datasets and compared with polynomial regression methods for reducing and reconstructing data. Experimental results show that data can be lossy compressed up to 75% for bandwidth utilization and 95.33% for storage space, with reconstruction accuracy close to 92%., This work was supported in part by the European Research Council (ERC) under the EU Horizon 2020 programme under Grant GA 639595, in part by the Spanish Ministry of Economy, Industry and Competitiveness under Grant TIN2015-65316-P and Grant IJCI2016-27485, in part by the Generalitat de Catalunya under Grant 2014-SGR-1051, in part by the University of the Punjab, Pakistan, and in part by the Qatar National Research Fund (a member of Qatar Foundation) under NPRP Grant # NPRP9-224-1-049., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Accelerating K-mer Frequency Counting with GPU and Non-Volatile Memory

  • Cadenelli, Nicola
  • Polo Cantero, José|||0000-0002-9917-1795
  • Carrera Pérez, David|||0000-0003-4898-3424
The emergence of Next Generation Sequencing (NGS) platforms has increased the throughput of genomic sequencing and in turn the amount of data that needs to be processed, requiring highly efficient computation for its analysis. In this context, modern architectures including accelerators and non-volatile memory are essential to enable the mass exploitation of these bioinformatics workloads. This paper presents a redesign of the main component of a state-of-the-art reference-free method for variant calling, SMUFIN, which has been adapted to make the most of GPUs and NVM devices. SMUFIN relies on counting the frequency of k-mers (substrings of length k) in DNA sequences, which also constitutes a well-known problem for many bioinformatics workloads, such as genome assembly. We propose techniques to improve the efficiency of k-mer counting and to scale-up workloads like SMUFIN that used to require 16 nodes of Marenostrum 3 to a single machine with a GPU and NVM drives. Results show that although the single machine is not able to improve the time to solution of 16 nodes, its CPU time is 7.5x shorter than the aggregate CPU time of the 16 nodes, with a reduction in energy consumption of 5.5x. © 2017 IEEE., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Multi-tenant Pub/Sub processing for real-time data streams

  • Villalba Navarro, Álvaro
  • Carrera Pérez, David|||0000-0003-4898-3424
Devices and sensors generate streams of data across a diversity of locations and protocols. That data usually reaches a central platform that is used to store and process the streams. Processing can be done in real time, with transformations and enrichment happening on-the-fly, but it can also happen after data is stored and organized in repositories. In the former case, stream processing technologies are required to operate on the data; in the latter batch analytics and queries are of common use. This paper introduces a runtime to dynamically construct data stream processing topologies based on user-supplied code. These dynamic topologies are built on-the-fly using a data subscription model defined by the applications that consume data. Each user-defined processing unit is called a Service Object. Every Service Object consumes input data streams and may produce output streams that others can consume. The subscription-based programing model enables multiple users to deploy their own data-processing services. The runtime does the dynamic forwarding of data and execution of Service Objects from different users. Data streams can originate in real-world devices or they can be the outputs of Service Objects. The runtime leverages Apache STORM for parallel data processing, that combined with dynamic user-code injection provides multi-tenant stream processing topologies. In this work we describe the runtime, its features and implementation details, as well as we include a performance evaluation of some of its core components., This work is partially supported by the European Research Council (ERC) un- der the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitivity (TIN2015-65316-P) and the Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Performance Characterization of Spark Workloads on Shared NUMA Systems

  • Baig, Shuja-ur-Rehman
  • Amaral, Marcelo
  • Polo, Jordà
  • Carrera, David
As the adoption of Big Data technologies becomes the norm in an increasing number of scenarios, there is also a growing need to optimize them for modern processors. Spark has gained momentum over the last few years among companies looking for high performance solutions that can scale out across different cluster sizes. At the same time, modern processors can be connected to large amounts of physical memory, in the range of up to few terabytes. This opens an enormous range of opportunities for runtimes and applications that aim to improve their performance by leveraging low latencies and high bandwidth provided by RAM. The result is that there are several examples today of applications that have started pushing the in-memory computing paradigm to accelerate tasks. To deliver such a large physical memory capacity, hardware vendors have leveraged Non-Uniform Memory Architectures (NUMA). This paper explores how Spark-based workloads are impacted by the effects of NUMA-placement decisions, how different Spark configurations result in changes in delivered performance, how the characteristics of the applications can be used to predict workload collocation conflicts, and how to improve performance by collocating workloads in scale-up nodes. We explore several workloads run on top of the IBM Power8 processor, and provide manual strategies that can leverage performance improvements up to 40% on Spark workloads when using smart processor-pinning and workload collocation strategies., This work is partially supported by the European Research Council (ERC) under the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitiveness (TIN2015-65316-P) and the Generalitat de Catalunya (2014-SGR-1051).
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Characterizing BigBench Queries, Hive, and Spark in Multi-cloud Environments

  • Poggi, Nicolas
  • Montero, Alejandro
  • Carrera, David
BigBench is the new standard (TPCx-BB) for benchmarking and testing Big Data systems. The TPCx-BB specification describes several business use cases—queries—which require a broad combination of data extraction techniques including SQL, Map/Reduce (M/R), user code (UDF), and Machine Learning to fulfill them. However, currently, there is no widespread knowledge of the different resource requirements and expected performance of each query, as is the case to more established benchmarks. Moreover, over the last year, the Spark framework and APIs have been evolving very rapidly, with major improvements in performance and the stable release of v2. It is our intent to compare the current state of Spark to Hive’s base implementation which can use the legacy M/R engine and Mahout or the current Tez and MLlib frameworks. At the same time, cloud providers currently offer convenient on-demand managed big data clusters (PaaS) with a pay-as-you-go model. In PaaS, analytical engines such as Hive and Spark come ready to use, with a general-purpose configuration and upgrade management. The study characterizes both the BigBench queries and the out-of-the-box performance of Spark and Hive versions in the cloud. At the same time, comparing popular PaaS offerings in terms of reliability, data scalability (1 GB to 10 TB), versions, and settings from Azure HDinsight, Amazon Web Services EMR, and Google Cloud Dataproc. The query characterization highlights the similarities and differences in Hive an Spark frameworks, and which queries are the most resource consuming according to CPU, memory, and I/O. Scalability results show how there is a need for configuration tuning in most cloud providers as data scale grows, especially with Sparks memory usage. These results can help practitioners to quickly test systems by picking a subset of the queries which stresses each of the categories. At the same time, results show how Hive and Spark compare and what performance can be expected of each in PaaS., This project has received funding from the European Research Council (ERC) under the European Unions Horizon 2020 research and innovation programme (grant agreement No. 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

ALOJA: A framework for benchmarking and predictive analytics in Hadoop deployments

  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Poggi Mastrokalo, Nicolas
  • Carrera Pérez, David|||0000-0003-4898-3424
  • Call Barreiro, Erin|||0000-0002-2770-1662
  • Reinauer, Rob
  • Green, Daron
This article presents the ALOJA project and its analytics tools, which leverages machine learning to interpret Big Data benchmark performance data and tuning. ALOJA is part of a long-term collaboration between BSC and Microsoft to automate the characterization of cost-effectiveness on Big Data deployments, currently focusing on Hadoop. Hadoop presents a complex run-time environment, where costs and performance depend on a large number of configuration choices. The ALOJA project has created an open, vendor-neutral repository, featuring over 40,000 Hadoop job executions and their performance details. The repository is accompanied by a test-bed and tools to deploy and evaluate the cost-effectiveness of different hardware configurations, parameters and Cloud services. Despite early success within ALOJA, a comprehensive study requires automation of modeling procedures to allow an analysis of large and resource-constrained search spaces. The predictive analytics extension, ALOJA-ML, provides an automated system allowing knowledge discovery by modeling environments from observed executions. The resulting models can forecast execution behaviors, predicting execution times for new configurations and hardware choices. That also enables model-based anomaly detection or efficient benchmark guidance by prioritizing executions. In addition, the community can benefit from ALOJA data-sets and framework to improve the design and deployment of Big Data applications., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). This work is partially supported by the Ministry of Economy of Spain under contracts TIN2012-34557 and 2014SGR1051., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Improving maritime traffic emission estimations on missing data with CRBMs

  • Gutiérrez Torre, Alberto|||0000-0002-5548-3359
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Buchaca Prats, David
  • Guevara Vilardell, Marc|||0000-0001-9727-8583
  • Soret, Albert|||0000-0002-1962-2972
  • Carrera Pérez, David|||0000-0003-4898-3424
© Elsevier 2020. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/, Maritime traffic emissions are a major concern to governments as they heavily impact the Air Quality in coastal cities. Ships use the Automatic Identification System (AIS) to continuously report position and speed among other features, and therefore this data is suitable to be used to estimate emissions, if it is combined with engine data. However, important ship features are often inaccurate or missing. State-of-the-art complex systems, like CALIOPE at the Barcelona Supercomputing Center, are used to model Air Quality. These systems can benefit from AIS based emission models as they are very precise in positioning the pollution. Unfortunately, these models are sensitive to missing or corrupted data, and therefore they need data curation techniques to significantly improve the estimation accuracy. In this work, we propose a methodology for treating ship data using Conditional Restricted Boltzmann Machines (CRBMs) plus machine learning methods to improve the quality of data passed to emission models that can also be applied to other GPS and time-series problems. Results show that we can improve the default methods proposed to cover missing data. In our results, we observed that using our method the models boosted their accuracy to detect otherwise undetectable emissions. In particular, we used a real data-set of AIS data, provided by the Spanish Port Authority, to estimate that thanks to our method, the model was able to detect 45% of additional emissions, representing 152 tonnes of pollutants per week in Barcelona and propose new features that may enhance emission modeling., This project hasreceived funding from the European Research Council (ERC) under the European Union Horizon 2020 research and innovation programme (grant agreement No 639595). This work is partially supported by the Ministry of Economy, Industry and Competitiveness of Spain under contracts TIN2015-65316-P,2014SGR1051, IJCI2016-27485, and Severo Ochoa Center of Excellence SEV-2015-0493-16-5., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Adaptive prediction models for data center resources utilization estimation

  • Baig, Shuja-ur-Rehman
  • Iqbal, Waheed
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Erradi, Abdelkarim
  • Carrera Pérez, David|||0000-0003-4898-3424
Accurate estimation of data center resource utilization is a challenging task due to multi-tenant co-hosted applications having dynamic and time-varying workloads. Accurate estimation of future resources utilization helps in better job scheduling, workload placement, capacity planning, proactive auto-scaling, and load balancing. The inaccurate estimation leads to either under or over-provisioning of data center resources. Most existing estimation methods are based on a single model that often does not appropriately estimate different workload scenarios. To address these problems, we propose a novel method to adaptively and automatically identify the most appropriate model to accurately estimate data center resources utilization. The proposed approach trains a classifier based on statistical features of historical resources usage to decide the appropriate prediction model to use for given resource utilization observations collected during a specific time interval. We evaluated our approach on real datasets and compared the results with multiple baseline methods. The experimental evaluation shows that the proposed approach outperforms the state-of-the-art approaches and delivers 6% to 27% improved resource utilization estimation accuracy compared to baseline methods., This work is partially supported by the European Research Council (ERC) under the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitiveness (TIN2015-65316-P and IJCI2016-27485), the Generalitat de Catalunya (2014-SGR-1051), and NPRP grant # NPRP9-224-1-049 from the Qatar National Research Fund (a member of Qatar Foundation) and University of the Punjab, Pakistan., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

The state of SQL-on-Hadoop in the cloud

  • Poggi, Nicolas
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Fenech, Thomas
  • Carrera Pérez, David|||0000-0003-4898-3424
  • Blakeley, Jose
  • Minhas, Umar F.
  • Vujic, Nikola
Managed Hadoop in the cloud, especially SQL-on-Hadoop, has been gaining attention recently. On Platform-as-a-Service (PaaS), analytical services like Hive and Spark come preconfigured for general-purpose and ready to use. Thus, giving companies a quick entry and on-demand deployment of ready SQL-like solutions for their big data needs. This study evaluates cloud services from an end-user perspective, comparing providers including: Microsoft Azure, Amazon Web Services, Google Cloud, and Rackspace. The study focuses on performance, readiness, scalability, and cost-effectiveness of the different solutions at entry/test level clusters sizes. Results are based on over 15,000 Hive queries derived from the industry standard TPC-H benchmark. The study is framed within the ALOJA research project, which features an open source benchmarking and analysis platform that has been recently extended to support SQL-on-Hadoop engines. The ALOJA Project aims to lower the total cost of ownership (TCO) of big data deployments and study their performance characteristics for optimization. The study benchmarks cloud providers across a diverse range instance types, and uses input data scales from 1GB to 1TB, in order to survey the popular entry-level PaaS SQL-on-Hadoop solutions, thereby establishing a common results-base upon which subsequent research can be carried out by the project. Initial results already show the main performance trends to both hardware and software configuration, pricing, similarities and architectural differences of the evaluated PaaS solutions. Whereas some providers focus on decoupling storage and computing resources while offering network-based elastic storage, others choose to keep the local processing model from Hadoop for high performance, but reducing flexibility. Results also show the importance of application-level tuning and how keeping up-to-date hardware and software stacks can influence performance even more than replicating the on-premises model in the cloud., This work is partially supported by the Microsoft Azure for Research program, the European Research Council (ERC) under the EUs Horizon 2020 programme (GA 639595), the Spanish Ministry of Education (TIN2015-65316-P), and the Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Constant-time sliding window framework with reduced memory footprint and efficient bulk evictions

  • Villalba Navarro, Álvaro
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Carrera Pérez, David|||0000-0003-4898-3424
The fast evolution of data analytics platforms has resulted in an increasing demand for real-time data stream processing. From Internet of Things applications to the monitoring of telemetry generated in large data centers, a common demand for currently emerging scenarios is the need to process vast amounts of data with low latencies, generally performing the analysis process as close to the data source as possible. Stream processing platforms are required to be malleable and absorb spikes generated by fluctuations of data generation rates. Data is usually produced as time series that have to be aggregated using multiple operators, being sliding windows one of the most common abstractions used to process data in real-time. To satisfy the above-mentioned demands, efficient stream processing techniques that aggregate data with minimal computational cost need to be developed. In this paper we present the Monoid Tree Aggregator general sliding window aggregation framework, which seamlessly combines the following features: amortized O(1) time complexity and a worst-case of O(log n) between insertions; it provides both a window aggregation mechanism and a window slide policy that are user programmable; the enforcement of the window sliding policy exhibits amortized O(1) computational cost for single evictions and supports bulk evictions with cost O(log n) ; and it requires a local memory space of O(log n) . The framework can compute aggregations over multiple data dimensions, and has been designed to support decoupling computation and data storage through the use of distributed Key-Value Stores to keep window elements and partial aggregations., This project is partially supported by the European Research Council (ERC), Spain under the European Unions Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015- 65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Database Integrated Analytics Using R: Initial Experiences with SQL-Server + R

  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Poggi, Nicolas
Most data scientists use nowadays functional or semi-functional languages like SQL, Scala or R to treat data, obtained directly from databases. Such process requires to fetch data, process it, then store again, and such process tends to be done outside the DB, in often complex data-flows. Recently, database service providers have decided to integrate "R-as-a-Service" in their DB solutions. The analytics engine is called directly from the SQL query tree, and results are returned as part of the same query. Here we show a first taste of such technology by testing the portability of our ALOJA-ML analytics framework, coded in R, to Microsoft SQL-Server 2016, one of the SQL+R solutions released recently. In this work we discuss some data-flow schemes for porting a local DB + analytics engine architecture towards Big Data, focusing specially on the new DB Integrated Analytics approach, and commenting the first experiences in usability and performance obtained from such new services and capabilities., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Topology-aware GPU scheduling for learning workloads in cloud environments

  • Amaral, Marcelo
  • Polo Bardés, Jordà
  • Carrera, David
  • Seelam, Seetharami
  • Steinder, Malgorzata
Recent advances in hardware, such as systems with multiple GPUs and their availability in the cloud, are enabling deep learning in various domains including health care, autonomous vehicles, and Internet of Things. Multi-GPU systems exhibit complex connectivity among GPUs and between GPUs and CPUs. Workload schedulers must consider hardware topology and workload communication requirements in order to allocate CPU and GPU resources for optimal execution time and improved utilization in shared cloud environments. This paper presents a new topology-aware workload placement strategy to schedule deep learning jobs on multi-GPU systems. The placement strategy is evaluated with a prototype on a Power8 machine with Tesla P100 cards, showing speedups of up to ≈1.30x compared to state-of-the-art strategies; the proposed algorithm achieves this result by allocating GPUs that satisfy workload requirements while preventing interference. Additionally, a large-scale simulation shows that the proposed strategy provides higher resource utilization and performance in cloud systems., This project is supported by the IBM/BSC Technology Center for Supercomputing collaboration agreement. It has also received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493). We thank our IBM Research colleagues Alaa Youssef and Asser Tantawi for the valuable discussions. We also thank SC17 committee member Blair Bethwaite of Monash University for his constructive feedback on the earlier drafts of this paper., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

When and How to Apply Statistics, Machine Learning and Deep Learning Techniques

  • Berral García, Josep Lluís|||0000-0003-3037-3580
Machine Learning has become 'commodity' in engineering and experimental sciences, as calculus and statistics did before. After the hype produced during the 00's, machine learning (statistical learning, neural networks, etc.) has become a solid and reliable set of techniques available to the general researcher population to be included in their common procedures, far from the mysticism surrounding this field when only ML experts could solve modeling and prediction problems using such novel algorithms. But while knowledge on this field has settled among professionals, novice ML users still have trouble to decide when determined techniques could and should be applied to solve a given problem, sometimes ending with over-complicated solutions for simplistic problems, or complex problems partially solved by simplistic methods. This tutorial wants to introduce the most common techniques on statistical learning and neural networks, towards showing the proper techniques for each given scenario., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Performance characterization of spark workloads on shared NUMA Systems

  • Baig, Shuja-ur-Rehman
  • Amaral, Marcelo
  • Polo Cantero, José|||0000-0002-9917-1795
  • Carrera Pérez, David|||0000-0003-4898-3424
As the adoption of Big Data technologies becomes the norm in an increasing number of scenarios, there is also a growing need to optimize them for modern processors. Spark has gained momentum over the last few years among companies looking for high performance solutions that can scale out across different cluster sizes. At the same time, modern processors can be connected to large amounts of physical memory, in the range of up to few terabytes. This opens an enormous range of opportunities for runtimes and applications that aim to improve their performance by leveraging low latencies and high bandwidth provided by RAM. The result is that there are several examples today of applications that have started pushing the in-memory computing paradigm to accelerate tasks. To deliver such a large physical memory capacity, hardware vendors have leveraged Non-Uniform Memory Architectures (NUMA). This paper explores how Spark-based workloads are impacted by the effects of NUMA-placement decisions, how different Spark configurations result in changes in delivered performance, how the characteristics of the applications can be used to predict workload collocation conflicts, and how to improve performance by collocating workloads in scale-up nodes. We explore several workloads run on top of the IBM Power8 processor, and provide manual strategies that can leverage performance improvements up to 40% on Spark workloads when using smart processor-pinning and workload collocation strategies., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Performance evaluation of microservices architectures using containers

  • Amaral, Marcelo
  • Polo Bardés, Jordà
  • Carrera Pérez, David|||0000-0003-4898-3424
  • Mohomed, Iqbal
  • Unuvar, Merve
  • Steinder, Malgorzata
Microservices architecture has started a new trend for application development for a number of reasons: (1) to reduce complexity by using tiny services; (2) to scale, remove and deploy parts of the system easily; (3) to improve flexibility to use different frameworks and tools; (4) to increase the overall scalability; and (5) to improve the resilience of the system. Containers have empowered the usage of microservices architectures by being lightweight, providing fast start-up times, and having a low overhead. Containers can be used to develop applications based on monolithic architectures where the whole system runs inside a single container or inside a microservices architecture where one or few processes run inside the containers. Two models can be used to implement a microservices architecture using containers: master-slave, or nested-container. The goal of this work is to compare the performance of CPU and network running benchmarks in the two aforementioned models of microservices architecture hence provide a benchmark analysis guidance for system designers., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Sequence-to-sequence models for workload interference prediction on batch processing datacenters

  • Buchaca Prats, David
  • Marcual Medina, Joan
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Carrera Pérez, David|||0000-0003-4898-3424
Co-scheduling of jobs in data centers is a challenging scenario where jobs can compete for resources, leading to severe slowdowns or failed executions. Efficient job placement on environments where resources are shared requires awareness on how jobs interfere during execution, to go far beyond ineffective resource overbooking techniques. Current techniques, most of which already involve machine learning and job modeling, are based on workload behavior summarization over time, rather than focusing on effective job requirements at each instant of the execution. In this work, we propose a methodology for modeling co-scheduling of jobs on data centers, based on their behavior towards resources and execution time and using sequence-to-sequence models based on recurrent neural networks. The goal is to forecast co-executed jobs footprint on resources throughout their execution time, from the profile shown by the individual jobs, in order to enhance resource manager and scheduler placement decisions. The methods presented herein are validated by using High Performance Computing benchmarks based on different frameworks (such as Hadoop and Spark) and applications (CPU bound, IO bound, machine learning, SQL queries...). Experiments show that the model can correctly identify the resource usage trends from previously seen and even unseen co-scheduled jobs., This work is supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement no. 639595); [Generalitat] de Catalunya under contract 2014SGR1051; the ICREA Academia program; and the BSC-CNS Severo Ochoa program (SEV-2015-0493); the Spanish Ministry of Economy under contract TIN2015-65316-P and the Generalitat., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

ALOJA-ML: a framework for automating characterization and knowledge discovery in Hadoop deployments

  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Poggi, Nicolas
  • Carrera Pérez, David|||0000-0003-4898-3424
  • Call Barreiro, Erin|||0000-0002-2770-1662
  • Reinauer, Rob
  • Green, Daron
This article presents ALOJA-Machine Learning (ALOJA-ML) an extension to the ALOJA project that uses machine learning techniques to interpret Hadoop benchmark performance data and performance tuning; here we detail the approach, efficacy of the model and initial results. The ALOJA-ML project is the latest phase of a long-term collaboration between BSC and Microsoft, to automate the characterization of cost-effectiveness on Big Data deployments, focusing on Hadoop. Hadoop presents a complex execution environment, where costs and performance depends on a large number of software (SW) configurations and on multiple hardware (HW) deployment choices. Recently the ALOJA project presented an open, vendor-neutral repository, featuring over 16.000 Hadoop executions. These results are accompanied by a test bed and tools to deploy and evaluate the cost-effectiveness of the different hardware configurations, parameter tunings, and Cloud services. Despite early success within ALOJA from expert-guided benchmarking, it became clear that a genuinely comprehensive study requires automation of modeling procedures to allow a systematic analysis of large and resource-constrained search spaces. ALOJA-ML provides such an automated system allowing knowledge discovery by modeling Hadoop executions from observed benchmarks across a broad set of configuration parameters. The resulting empirically-derived performance models can be used to forecast execution behavior of various workloads; they allow a-priori prediction of the execution times for new configurations and HW choices and they offer a route to model-based anomaly detection. In addition, these models can guide the benchmarking exploration efficiently, by automatically prioritizing candidate future benchmark tests. Insights from ALOJA-ML's models can be used to reduce the operational time on clusters, speed-up the data acquisition and knowledge discovery process, and importantly, reduce running costs. In addition to learning from the methodology presented in this work, the community can benefit in general from ALOJA data-sets, framework, and derived insights to improve the design and deployment of Big Data applications., This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 re- search and innovation programme (grant agreement No 639595). This work is partially supported by the Ministry of Economy of Spain under contracts TIN2012-34557 and 2014SGR105, Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Constant-Time Sliding Window Framework with Reduced Memory Footprint and Efficient Bulk Evictions

  • Villalba, Álvaro
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Carrera, David
The fast evolution of data analytics platforms has resulted in an increasing demand for real-time data stream processing. From Internet of Things applications to the monitoring of telemetry generated in large data centers, a common demand for currently emerging scenarios is the need to process vast amounts of data with low latencies, generally performing the analysis process as close to the data source as possible. Stream processing platforms are required to be malleable and absorb spikes generated by fluctuations of data generation rates. Data is usually produced as time series that have to be aggregated using multiple operators, being sliding windows one of the most common abstractions used to process data in real-time. To satisfy the above-mentioned demands, efficient stream processing techniques that aggregate data with minimal computational cost need to be developed. In this paper we present the Monoid Tree Aggregator general sliding window aggregation framework, which seamlessly combines the following features: amortized O(1) time complexity and a worst-case of O(logn) between insertions; it provides both a window aggregation mechanism and a window slide policy that are user programmable; the enforcement of the window sliding policy exhibits amortized O(1) computational cost for single evictions and supports bulk evictions with cost O(logn) ; and it requires a local memory space of O(logn) . The framework can compute aggregations over multiple data dimensions, and has been designed to support decoupling computation and data storage through the use of distributed Key-Value Stores to keep window elements and partial aggregations., This project is partially supported by the European ResearchCouncil (ERC), Spain under the European Unions Horizon2020 research and innovation programme (grant agreementNo 639595). It is also partially supported by the Ministryof Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051,by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

You only run once: Spark auto-tuning from a single run

  • Buchaca Prats, David
  • Albuquerque Portella, Felipe|||0000-0003-1245-7465
  • Costa, Carlos H. A.
  • Berral García, Josep Lluís|||0000-0003-3037-3580
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., This work is supported by the European Research Council under the European Union’s Horizon 2020 research and innovation programme (grant agreement no 639595); the Spanish Ministry of Economy, under contract TIN2015- 65316-P, and the Generalitat de Catalunya under contract 2014SGR1051; the ICREA Academia program; the BSC-CNS Severo Ochoa program (SEV-2015-0493); and by Petroleo Brasileiro S. A. (PETROBRAS)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Enabling genomics pipelines in commodity personal computers with flash storage

  • Cadenelli, Nicola
  • Jun, Sang-Woo
  • Polo Bardés, Jorda
  • Wright, Andrew J.
  • Carrera Pérez, David|||0000-0003-4898-3424
  • Mithal, Arvind
Analysis of a patient’s genomics data is the first step toward precision medicine. Such analyses are performed on expensive enterprise-class server machines because input data sets are large, and the intermediate data structures are even larger (TB-size) and require random accesses. We present a general method to perform a specific genomics problem, mutation detection, on a cheap commodity personal computer (PC) with a small amount of DRAM. We construct and access large histograms of k-mers efficiently on external storage (SSDs) and apply our technique to a state-of-the-art referencefree genomics algorithm, SMUFIN, to create SMUFIN-F. We show that on two PCs, SMUFIN-F can achieve the same throughput at only one third (36%) the hardware cost and half (45%) the energy compared to SMUFIN on an enterprise-class server. To the best of our knowledge, SMUFIN-F is the first reference-free system that can detect somatic mutations on commodity PCs for whole human genomes. We believe our technique should apply to other k-mer or n-gram-based algorithms., This work was supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreements No. 639595); the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051; the ICREA Academia program; the BSC-CNS Severo Ochoa program (SEV-2015-0493); MIT via the NSF grant (CCF1725303); and travel support was provided by the MIT-Spain la Caixa Foundation Seed Fund., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

A highly parameterizable framework for Conditional Restricted Boltzmann Machine based workloads accelerated with FPGAs and OpenCL

  • Jaksic, Zoran
  • Cadenelli, Nicola
  • Buchaca Prats, David
  • Polo Bardés, Jordà
  • Berral García, Josep Lluís|||0000-0003-3037-3580
  • Carrera Pérez, David|||0000-0003-4898-3424
© 2020 Elsevier. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/, Conditional Restricted Boltzmann Machine (CRBM) is a promising candidate for a multidimensional system modeling that can learn a probability distribution over a set of data. It is a specific type of an artificial neural network with one input (visible) and one output (hidden) layer. Recently published works demonstrate that CRBM is a suitable mechanism for modeling multidimensional time series such as human motion, workload characterization, city traffic analysis. The process of learning and inference of these systems relies on linear algebra functions like matrix–matrix multiplication, and for higher data sets, they are very compute-intensive. In this paper, we present a configurable framework for CRBM based workloads for arbitrary large models. We show how to accelerate the learning process of CRBM with FPGAs and OpenCL, and we conduct an extensive scalability study for different model sizes and system configurations. We show significant improvement in performance/Watt for large models and batch sizes (from 1.51x up to 5.71x depending on the host configuration) when we use FPGA and OpenCL for the acceleration, and limited benefits for small models comparing to the state-of-the-art CPU solution., This work was supported by the European Research Council(ERC) under the European Union’s Horizon 2020 research andinnovation programme (grant agreements No 639595); the Min-istry of Economy of Spain under contract TIN2015-65316-P andGeneralitat de Catalunya, Spain under contract 2014SGR1051;the ICREA, Spain Academia program; the BSC-CNS Severo Ochoaprogram, Spain (SEV-2015-0493) and Intel Corporation, UnitedStates, Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Considerations in using OpenCL on GPUs and FPGAs for throughput-oriented genomics workloads

  • Cadenelli, Nicola
  • Jaksic, Zoran
  • Polo Bardés, Jordà
  • Carrera, David
The recent upsurge in the available amount of health data and the advances in next-generation sequencing are setting the ground for the long-awaited precision medicine. To process this deluge of data, bioinformatics workloads are becoming more complex and more computationally demanding. For this reasons they have been extended to support different computing architectures, such as GPUs and FPGAs, to leverage the form of parallelism typical of each of such architectures. The paper describes how a genomic workload such as k-mer frequency counting that takes advantage of a GPU can be offloaded to one or even more FPGAs. Moreover, it performs a comprehensive analysis of the FPGA acceleration comparing its performance to a non-accelerated configuration and when using a GPU. Lastly, the paper focuses on how, when using accelerators with a throughput-oriented workload, one should also take into consideration both kernel execution time and how well each accelerator board overlaps kernels and PCIe transferred. Results show that acceleration with two FPGAs can improve both time- and energy-to-solution for the entire accelerated part by a factor of 1.32x. Per contra, acceleration with one GPU delivers an improvement of 1.77x in time-to-solution but of a lower 1.49x in energy-to-solution due to persistently higher power consumption. The paper also evaluates how future FPGA boards with components (i.e., off-chip memory and PCIe) on par with those of the GPU board could provide an energy-efficient alternative to GPUs., This work was supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement s No 639595); the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051; the ICREA, Spain Academia program; and the BSC-CNS Severo Ochoa, Spain program (SEV-2015-0493)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Disaggregating non-volatile memory for throughput-oriented genomics workloads

  • Call Barreiro, Erin|||0000-0002-2770-1662
  • Polo Bardés, Jordà
  • Carrera Pérez, David|||0000-0003-4898-3424
  • Guim Bernat, Francesc
  • Sen, Sujoy
Massive exploitation of next-generation sequencing technologies requires dealing with both: huge amounts of data and complex bioinformatics pipelines. Computing architectures have evolved to deal with these problems, enabling approaches that were unfeasible years ago: accelerators and Non-Volatile Memories (NVM) are becoming widely used to enhance the most demanding workloads. However, bioinformatics workloads are usually part of bigger pipelines with different and dynamic needs in terms of resources. The introduction of Software Defined Infrastructures (SDI) for data centers provides roots to dramatically increase the efficiency in the management of infrastructures. SDI enables new ways to structure hardware resources through disaggregation, and provides new hardware composability and sharing mechanisms to deploy workloads in more flexible ways. In this paper we study a state-of-the-art genomics application, SMUFIN, aiming to address the challenges of future HPC facilities., This work is partially supported by the European Research Council (ERC) under the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitivity (TIN2015-65316-P) and the Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Contribución de congreso (ConferenceOutput)

Constant-time approximate sliding window framework with error control

  • Villalba Navarro, Álvaro
  • Carrera Pérez, David|||0000-0003-4898-3424
Stream Processing is a crucial element for the Edge Computing paradigm, in which large amount of devices generate data at the edge of the network. This data needs to be aggregated and processed on-the-move across different layers before reaching the Cloud. Therefore, defining Stream Processing services that adapt to different levels of resource availability is of paramount importance. In this context, Stream Processing frameworks need to combine efficient algorithms with low computational complexity to manage sliding windows, with the ability to adjust resource demands for different deployment scenarios, from very low capacity edge devices to virtually unlimited Cloud platforms. The Approximate Computing paradigm provides improved performance and adaptive resource demands in data analytics, at the price of introducing some level of inaccuracy that can be calculated. In this paper we present the Approximate and Amortized Monoid Tree Aggregator (A 2 MTA). It is, to our knowledge, the first general purpose sliding window programable framework that combines constant-time aggregations with error bounded approximate computing techniques. It is very suitable for adverse stream processing environments, such as resource scarce multi-tenant edge computing. The framework can compute aggregations over multiple data dimensions, setting error bounds on any of them, and has been designed to support decoupling computation and data storage through the use of distributed Key-Value Stores to keep window elements and partial aggregations., This project is partially supported by the European Research Council (ERC), Spain under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Accelerating K-mer Frequency Counting with GPU and Non-Volatile Memory

  • Cadenelli, Nicola
  • Polo Cantero, José
  • Carrera Pérez, David
The emergence of Next Generation Sequencing (NGS) platforms has increased the throughput of genomic sequencing and in turn the amount of data that needs to be processed, requiring highly efficient computation for its analysis. In this context, modern architectures including accelerators and non-volatile memory are essential to enable the mass exploitation of these bioinformatics workloads. This paper presents a redesign of the main component of a state-of-the-art reference-free method for variant calling, SMUFIN, which has been adapted to make the most of GPUs and NVM devices. SMUFIN relies on counting the frequency of k-mers (substrings of length k) in DNA sequences, which also constitutes a well-known problem for many bioinformatics workloads, such as genome assembly. We propose techniques to improve the efficiency of k-mer counting and to scale-up workloads like SMUFIN that used to require 16 nodes of Marenostrum 3 to a single machine with a GPU and NVM drives. Results show that although the single machine is not able to improve the time to solution of 16 nodes, its CPU time is 7.5x shorter than the aggregate CPU time of the 16 nodes, with a reduction in energy consumption of 5.5x. © 2017 IEEE., Peer Reviewed, Preprint
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Adaptive prediction models for data center resources utilization estimation

  • Baig, Shuja-ur-Rehman
  • Iqbal, Waheed
  • Berral García, Josep Lluís
  • Erradi, Abdelkarim
  • Carrera Pérez, David
Accurate estimation of data center resource utilization is a challenging task due to multi-tenant co-hosted applications having dynamic and time-varying workloads. Accurate estimation of future resources utilization helps in better job scheduling, workload placement, capacity planning, proactive auto-scaling, and load balancing. The inaccurate estimation leads to either under or over-provisioning of data center resources. Most existing estimation methods are based on a single model that often does not appropriately estimate different workload scenarios. To address these problems, we propose a novel method to adaptively and automatically identify the most appropriate model to accurately estimate data center resources utilization. The proposed approach trains a classifier based on statistical features of historical resources usage to decide the appropriate prediction model to use for given resource utilization observations collected during a specific time interval. We evaluated our approach on real datasets and compared the results with multiple baseline methods. The experimental evaluation shows that the proposed approach outperforms the state-of-the-art approaches and delivers 6% to 27% improved resource utilization estimation accuracy compared to baseline methods., This work is partially supported by the European Research Council (ERC) under the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitiveness (TIN2015-65316-P and IJCI2016-27485), the Generalitat de Catalunya (2014-SGR-1051), and NPRP grant # NPRP9-224-1-049 from the Qatar National Research Fund (a member of Qatar Foundation) and University of the Punjab, Pakistan., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

A Quick View on Current Techniques and Machine Learning Algorithms for Big Data Analytics

  • Berral García, Josep Lluís
Big-data is an excellent source of knowledge and information from our systems and clients, but dealing with such amount of data requires automation, and this brings us to data mining and machine leaming techniques. In the ICT sector, as in many other sectors of research and industry, platforms and tools are being served and developed in order to help professionals to treat their data and leam from it automatically ; most of those platforms coming from big companies like Google or Microsoft , or from incubators at the Apache Foundation. This brief review explains the basics of machine learning with sorne ICT examples, and enurnerates sorne (but not all) of the most used tools for analyzing and modelling big-data., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

DRMaestro: orchestrating disaggregated resources on virtualized data-centers

  • Carneiro do Amaral, Marcelo
  • Polo Bardés, Jorda
  • Carrera Pérez, David
  • González, Nelson
  • Yang, Chih-Chieh
  • Morari, Alessandro
  • D'Amora, Bruce
  • Youssef, Alaa
  • Steinder, Malgorzata
Modern applications demand resources at an unprecedented level. In this sense, data-centers are required to scale efficiently to cope with such demand. Resource disaggregation has the potential to improve resource-efficiency by allowing the deployment of workloads in more flexible ways. Therefore, the industry is shifting towards disaggregated architectures, which enables new ways to structure hardware resources in data centers. However, determining the best performing resource provisioning is a complicated task. The optimality of resource allocation in a disaggregated data center depends on its topology and the workload collocation. This paper presents DRMaestro, a framework to orchestrate disaggregated resources transparently from the applications. DRMaestro uses a novel flow-network model to determine the optimal placement in multiple phases while employing best-efforts on preventing workload performance interference. We first evaluate the impact of disaggregation regarding the additional network requirements under higher network load. The results show that for some applications the impact is minimal, but other ones can suffer up to 80% slowdown in the data transfer part. After that, we evaluate DRMaestro via a real prototype on Kubernetes and a trace-driven simulation. The results show that DRMaestro can reduce the total job makespan with a speedup of up to ≈1.20x and decrease the QoS violation up to ≈2.64x comparing with another orchestrator that does not support resource disaggregation., This project is supported by the IBM/BSC Technology Center for Supercomputing collaboration agreement. It has also received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Database integrated analytics using R : initial experiences with SQL-Server + R

  • Berral García, Josep Lluís
  • Poggi, Nicolas
© 2016 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Most data scientists use nowadays functional or semi-functional languages like SQL, Scala or R to treat data, obtained directly from databases. Such process requires to fetch data, process it, then store again, and such process tends to be done outside the DB, in often complex data-flows. Recently, database service providers have decided to integrate “R-as-a-Service” in their DB solutions. The analytics engine is called directly from the SQL query tree, and results are returned as part of the same query. Here we show a first taste of such technology by testing the portability of our ALOJA-ML analytics framework, coded in R, to Microsoft SQL-Server 2016, one of the SQL+R solutions released recently. In this work we discuss some data-flow schemes for porting a local DB + analytics engine architecture towards Big Data, focusing specially on the new DB Integrated Analytics approach, and commenting the first experiences in usability and performance obtained from such new services and capabilities., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

A highly parameterizable framework for Conditional Restricted Boltzmann Machine based workloads accelerated with FPGAs and OpenCL

  • Jaksic, Zoran
  • Cadenelli, Nicola
  • Buchaca Prats, David
  • Polo Bardés, Jordà
  • Berral García, Josep Lluís
  • Carrera Pérez, David
© 2020 Elsevier. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/, Conditional Restricted Boltzmann Machine (CRBM) is a promising candidate for a multidimensional system modeling that can learn a probability distribution over a set of data. It is a specific type of an artificial neural network with one input (visible) and one output (hidden) layer. Recently published works demonstrate that CRBM is a suitable mechanism for modeling multidimensional time series such as human motion, workload characterization, city traffic analysis. The process of learning and inference of these systems relies on linear algebra functions like matrix–matrix multiplication, and for higher data sets, they are very compute-intensive. In this paper, we present a configurable framework for CRBM based workloads for arbitrary large models. We show how to accelerate the learning process of CRBM with FPGAs and OpenCL, and we conduct an extensive scalability study for different model sizes and system configurations. We show significant improvement in performance/Watt for large models and batch sizes (from 1.51x up to 5.71x depending on the host configuration) when we use FPGA and OpenCL for the acceleration, and limited benefits for small models comparing to the state-of-the-art CPU solution., This work was supported by the European Research Council(ERC) under the European Union’s Horizon 2020 research andinnovation programme (grant agreements No 639595); the Min-istry of Economy of Spain under contract TIN2015-65316-P andGeneralitat de Catalunya, Spain under contract 2014SGR1051;the ICREA, Spain Academia program; the BSC-CNS Severo Ochoaprogram, Spain (SEV-2015-0493) and Intel Corporation, UnitedStates, Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Automatic generation of workload profiles using unsupervised learning pipelines

  • Buchaca Prats, David
  • Berral García, Josep Lluís
  • Carrera Pérez, David
The complexity of resource usage and power consumption on cloud-based applications makes the understanding of application behavior through expert examination difficult. The difficulty increases when applications are seen as “black boxes”, where only external monitoring can be retrieved. Furthermore, given the different amount of scenarios and applications, automation is required. Here we examine and model application behavior by finding behavior phases. We use Conditional Restricted Boltzmann Machines (CRBM) to model time-series containing resources traces measurements like CPU, Memory and IO. CRBMs can be used to map a given given historic window of trace behaviour into a single vector. This low dimensional and time-aware vector can be passed through clustering methods, from simplistic ones like k-means to more complex ones like those based on Hidden Markov Models (HMM). We use these methods to find phases of similar behaviour in the workloads. Our experimental evaluation shows that the proposed method is able to identify different phases of resource consumption across different workloads. We show that the distinct phases contain specific resource patterns that distinguish them., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Constant-time approximate sliding window framework with error control

  • Villalba Navarro, Álvaro
  • Carrera Pérez, David
Stream Processing is a crucial element for the Edge Computing paradigm, in which large amount of devices generate data at the edge of the network. This data needs to be aggregated and processed on-the-move across different layers before reaching the Cloud. Therefore, defining Stream Processing services that adapt to different levels of resource availability is of paramount importance. In this context, Stream Processing frameworks need to combine efficient algorithms with low computational complexity to manage sliding windows, with the ability to adjust resource demands for different deployment scenarios, from very low capacity edge devices to virtually unlimited Cloud platforms. The Approximate Computing paradigm provides improved performance and adaptive resource demands in data analytics, at the price of introducing some level of inaccuracy that can be calculated. In this paper we present the Approximate and Amortized Monoid Tree Aggregator (A 2 MTA). It is, to our knowledge, the first general purpose sliding window programable framework that combines constant-time aggregations with error bounded approximate computing techniques. It is very suitable for adverse stream processing environments, such as resource scarce multi-tenant edge computing. The framework can compute aggregations over multiple data dimensions, setting error bounds on any of them, and has been designed to support decoupling computation and data storage through the use of distributed Key-Value Stores to keep window elements and partial aggregations., This project is partially supported by the European Research Council (ERC), Spain under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Enabling genomics pipelines in commodity personal computers with flash storage

  • Cadenelli, Nicola
  • Jun, Sang-Woo
  • Polo Bardés, Jorda
  • Wright, Andrew J.
  • Carrera Pérez, David
  • Mithal, Arvind
Analysis of a patient’s genomics data is the first step toward precision medicine. Such analyses are performed on expensive enterprise-class server machines because input data sets are large, and the intermediate data structures are even larger (TB-size) and require random accesses. We present a general method to perform a specific genomics problem, mutation detection, on a cheap commodity personal computer (PC) with a small amount of DRAM. We construct and access large histograms of k-mers efficiently on external storage (SSDs) and apply our technique to a state-of-the-art referencefree genomics algorithm, SMUFIN, to create SMUFIN-F. We show that on two PCs, SMUFIN-F can achieve the same throughput at only one third (36%) the hardware cost and half (45%) the energy compared to SMUFIN on an enterprise-class server. To the best of our knowledge, SMUFIN-F is the first reference-free system that can detect somatic mutations on commodity PCs for whole human genomes. We believe our technique should apply to other k-mer or n-gram-based algorithms., This work was supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreements No. 639595); the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051; the ICREA Academia program; the BSC-CNS Severo Ochoa program (SEV-2015-0493); MIT via the NSF grant (CCF1725303); and travel support was provided by the MIT-Spain la Caixa Foundation Seed Fund., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Burst-aware predictive autoscaling for containerized microservices

  • Abdullah, Muhammad
  • Iqbal, Waheed
  • Berral García, Josep Lluís
  • Polo Bardés, Jorda
  • Carrera Pérez, David
Autoscaling methods are used for cloud-hosted applications to dynamically scale the allocated resources for guaranteeing Quality-of-Service (QoS). The public-facing application serves dynamic workloads, which contain bursts and pose challenges for autoscaling methods to ensure application performance. Existing State-of-the-art autoscaling methods are burst-oblivious to determine and provision the appropriate resources. For dynamic workloads, it is hard to detect and handle bursts online for maintaining application performance. In this article, we propose a novel burst-aware autoscaling method which detects burst in dynamic workloads using workload forecasting, resource prediction, and scaling decision making while minimizing response time service-level objectives (SLO) violations. We evaluated our approach through a trace-driven simulation, using multiple synthetic and realistic bursty workloads for containerized microservices, improving performance when comparing against existing state-of-the-art autoscaling methods. Such experiments show an increase of × 1.09 in total processed requests, a reduction of × 5.17 for SLO violations, and an increase of × 0.767 cost as compared to the baseline method., This work was partially supported by the European Research Council (ERC) under the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitiveness (TIN2015-65316-P and IJCI2016-27485) and the Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

The state of SQL-on-Hadoop in the cloud

  • Poggi, Nicolas
  • Berral García, Josep Lluís
  • Fenech, Thomas
  • Carrera Pérez, David
  • Blakeley, Jose
  • Minhas, Umar F.
  • Vujic, Nikola
Managed Hadoop in the cloud, especially SQL-on-Hadoop, has been gaining attention recently. On Platform-as-a-Service (PaaS), analytical services like Hive and Spark come preconfigured for general-purpose and ready to use. Thus, giving companies a quick entry and on-demand deployment of ready SQL-like solutions for their big data needs. This study evaluates cloud services from an end-user perspective, comparing providers including: Microsoft Azure, Amazon Web Services, Google Cloud, and Rackspace. The study focuses on performance, readiness, scalability, and cost-effectiveness of the different solutions at entry/test level clusters sizes. Results are based on over 15,000 Hive queries derived from the industry standard TPC-H benchmark. The study is framed within the ALOJA research project, which features an open source benchmarking and analysis platform that has been recently extended to support SQL-on-Hadoop engines. The ALOJA Project aims to lower the total cost of ownership (TCO) of big data deployments and study their performance characteristics for optimization. The study benchmarks cloud providers across a diverse range instance types, and uses input data scales from 1GB to 1TB, in order to survey the popular entry-level PaaS SQL-on-Hadoop solutions, thereby establishing a common results-base upon which subsequent research can be carried out by the project. Initial results already show the main performance trends to both hardware and software configuration, pricing, similarities and architectural differences of the evaluated PaaS solutions. Whereas some providers focus on decoupling storage and computing resources while offering network-based elastic storage, others choose to keep the local processing model from Hadoop for high performance, but reducing flexibility. Results also show the importance of application-level tuning and how keeping up-to-date hardware and software stacks can influence performance even more than replicating the on-premises model in the cloud., This work is partially supported by the Microsoft Azure for Research program, the European Research Council (ERC) under the EUs Horizon 2020 programme (GA 639595), the Spanish Ministry of Education (TIN2015-65316-P), and the Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

A resilient and distributed near real-time traffic forecasting application for Fog computing environments

  • Pérez, Juan Luis
  • Gutiérrez Torre, Alberto
  • Berral García, Josep Lluís
  • Carrera Pérez, David
In this paper we propose an architecture for a city-wide traffic modeling and prediction service based on the Fog Computing paradigm. The work assumes an scenario in which a number of distributed antennas receive data generated by vehicles across the city. In the Fog nodes data is collected, processed in local and intermediate nodes, and finally forwarded to a central Cloud location for further analysis. We propose a combination of a data distribution algorithm, resilient to back-haul connectivity issues, and a traffic modeling approach based on deep learning techniques to provide distributed traffic forecasting capabilities. In our experiments, we leverage real traffic logs from one week of Floating Car Data (FCD) generated in the city of Barcelona by a road-assistance service fleet comprising thousands of vehicles. FCD was processed across several simulated conditions, ranging from scenarios in which no connectivity failures occurred in the Fog nodes, to situations with long and frequent connectivity outage periods. For each scenario, the resilience and accuracy of both the data distribution algorithm, and the learning methods were analyzed. Results show that the data distribution process running in the Fog nodes is resilient to back-haul connectivity issues and is able to deliver data to the Cloud location even in presence of severe connectivity problems. Additionally, the proposed traffic modeling and forecasting method exhibits better behavior when run distributed in the Fog instead of centralized in the Cloud, especially when connectivity issues occur that force data to be delivered out of order to the Cloud., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

When and How to Apply Statistics, Machine Learning and Deep Learning Techniques

  • Berral García, Josep Lluís
Machine Learning has become 'commodity' in engineering and experimental sciences, as calculus and statistics did before. After the hype produced during the 00's, machine learning (statistical learning, neural networks, etc.) has become a solid and reliable set of techniques available to the general researcher population to be included in their common procedures, far from the mysticism surrounding this field when only ML experts could solve modeling and prediction problems using such novel algorithms. But while knowledge on this field has settled among professionals, novice ML users still have trouble to decide when determined techniques could and should be applied to solve a given problem, sometimes ending with over-complicated solutions for simplistic problems, or complex problems partially solved by simplistic methods. This tutorial wants to introduce the most common techniques on statistical learning and neural networks, towards showing the proper techniques for each given scenario., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Database Integrated Analytics Using R: Initial Experiences with SQL-Server + R

  • Berral García, Josep Lluís
  • Poggi, Nicolas
Most data scientists use nowadays functional or semi-functional languages like SQL, Scala or R to treat data, obtained directly from databases. Such process requires to fetch data, process it, then store again, and such process tends to be done outside the DB, in often complex data-flows. Recently, database service providers have decided to integrate "R-as-a-Service" in their DB solutions. The analytics engine is called directly from the SQL query tree, and results are returned as part of the same query. Here we show a first taste of such technology by testing the portability of our ALOJA-ML analytics framework, coded in R, to Microsoft SQL-Server 2016, one of the SQL+R solutions released recently. In this work we discuss some data-flow schemes for porting a local DB + analytics engine architecture towards Big Data, focusing specially on the new DB Integrated Analytics approach, and commenting the first experiences in usability and performance obtained from such new services and capabilities., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

ALOJA-ML: a framework for automating characterization and knowledge discovery in Hadoop deployments

  • Berral García, Josep Lluís
  • Poggi, Nicolas
  • Carrera Pérez, David
  • Call Barreiro, Erin
  • Reinauer, Rob
  • Green, Daron
This article presents ALOJA-Machine Learning (ALOJA-ML) an extension to the ALOJA project that uses machine learning techniques to interpret Hadoop benchmark performance data and performance tuning; here we detail the approach, efficacy of the model and initial results. The ALOJA-ML project is the latest phase of a long-term collaboration between BSC and Microsoft, to automate the characterization of cost-effectiveness on Big Data deployments, focusing on Hadoop. Hadoop presents a complex execution environment, where costs and performance depends on a large number of software (SW) configurations and on multiple hardware (HW) deployment choices. Recently the ALOJA project presented an open, vendor-neutral repository, featuring over 16.000 Hadoop executions. These results are accompanied by a test bed and tools to deploy and evaluate the cost-effectiveness of the different hardware configurations, parameter tunings, and Cloud services. Despite early success within ALOJA from expert-guided benchmarking, it became clear that a genuinely comprehensive study requires automation of modeling procedures to allow a systematic analysis of large and resource-constrained search spaces. ALOJA-ML provides such an automated system allowing knowledge discovery by modeling Hadoop executions from observed benchmarks across a broad set of configuration parameters. The resulting empirically-derived performance models can be used to forecast execution behavior of various workloads; they allow a-priori prediction of the execution times for new configurations and HW choices and they offer a route to model-based anomaly detection. In addition, these models can guide the benchmarking exploration efficiently, by automatically prioritizing candidate future benchmark tests. Insights from ALOJA-ML's models can be used to reduce the operational time on clusters, speed-up the data acquisition and knowledge discovery process, and importantly, reduce running costs. In addition to learning from the methodology presented in this work, the community can benefit in general from ALOJA data-sets, framework, and derived insights to improve the design and deployment of Big Data applications., This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 re- search and innovation programme (grant agreement No 639595). This work is partially supported by the Ministry of Economy of Spain under contracts TIN2012-34557 and 2014SGR105, Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Accelerating K-mer Frequency Counting with GPU and Non-Volatile Memory

  • Cadenelli, Nicola
  • Polo Bardés, Jordà
  • Carrera, David
The emergence of Next Generation Sequencing (NGS) platforms has increased the throughput of genomic sequencing and in turn the amount of data that needs to be processed, requiring highly efficient computation for its analysis. In this context, modern architectures including accelerators and non-volatile memory are essential to enable the mass exploitation of these bioinformatics workloads. This paper presents a redesign of the main component of a state-of-the-art reference-free method for variant calling, SMUFIN, which has been adapted to make the most of GPUs and NVM devices. SMUFIN relies on counting the frequency of k-mers (substrings of length k) in DNA sequences, which also constitutes a well-known problem for many bioinformatics workloads, such as genome assembly. We propose techniques to improve the efficiency of k-mer counting and to scale-up workloads like SMUFIN that used to require 16 nodes of Marenostrum 3 to a single machine with a GPU and NVM drives. Results show that although the single machine is not able to improve the time to solution of 16 nodes, its CPU time is 7.5x shorter than the aggregate CPU time of the 16 nodes, with a reduction in energy consumption of 5.5x., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493). We are also grateful to SandDisk for lending the FusionIO cards and to Nvidia who donated the Tesla K40c., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

A resilient and distributed near real-time traffic forecasting application for Fog computing environments

  • Pérez, Juan L.
  • Gutierrez-Torre, Alberto
  • Berral García, Josep Lluís
  • Carrera, David
In this paper we propose an architecture for a city-wide traffic modeling and prediction service based on the Fog Computing paradigm. The work assumes an scenario in which a number of distributed antennas receive data generated by vehicles across the city. In the Fog nodes data is collected, processed in local and intermediate nodes, and finally forwarded to a central Cloud location for further analysis. We propose a combination of a data distribution algorithm, resilient to back-haul connectivity issues, and a traffic modeling approach based on deep learning techniques to provide distributed traffic forecasting capabilities. In our experiments, we leverage real traffic logs from one week of Floating Car Data (FCD) generated in the city of Barcelona by a road-assistance service fleet comprising thousands of vehicles. FCD was processed across several simulated conditions, ranging from scenarios in which no connectivity failures occurred in the Fog nodes, to situations with long and frequent connectivity outage periods. For each scenario, the resilience and accuracy of both the data distribution algorithm, and the learning methods were analyzed. Results show that the data distribution process running in the Fog nodes is resilient to back-haul connectivity issues and is able to deliver data to the Cloud location even in presence of severe connectivity problems. Additionally, the proposed traffic modeling and forecasting method exhibits better behavior when run distributed in the Fog instead of centralized in the Cloud, especially when connectivity issues occur that force data to be delivered out of order to the Cloud., This project is partially supported by the European Research Council (ERC), Spain under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493). The authors gratefully acknowledge the Reial Automvil Club de Catalunya (RACC) for the dataset of Floating Car Data provided., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

You only run once: Spark auto-tuning from a single run

  • Buchaca Prats, David
  • Albuquerque Portella, Felipe
  • Costa, Carlos H. A.
  • Berral García, Josep Lluís
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., This work is supported by the European Research Council under the European Union’s Horizon 2020 research and innovation programme (grant agreement no 639595); the Spanish Ministry of Economy, under contract TIN2015- 65316-P, and the Generalitat de Catalunya under contract 2014SGR1051; the ICREA Academia program; the BSC-CNS Severo Ochoa program (SEV-2015-0493); and by Petroleo Brasileiro S. A. (PETROBRAS)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Characterizing BigBench Queries, Hive, and Spark in Multi-cloud Environments

  • Poggi, Nicolas
  • Montero, Alejandro
  • Carrera, David
BigBench is the new standard (TPCx-BB) for benchmarking and testing Big Data systems. The TPCx-BB specification describes several business use cases—queries—which require a broad combination of data extraction techniques including SQL, Map/Reduce (M/R), user code (UDF), and Machine Learning to fulfill them. However, currently, there is no widespread knowledge of the different resource requirements and expected performance of each query, as is the case to more established benchmarks. Moreover, over the last year, the Spark framework and APIs have been evolving very rapidly, with major improvements in performance and the stable release of v2. It is our intent to compare the current state of Spark to Hive’s base implementation which can use the legacy M/R engine and Mahout or the current Tez and MLlib frameworks. At the same time, cloud providers currently offer convenient on-demand managed big data clusters (PaaS) with a pay-as-you-go model. In PaaS, analytical engines such as Hive and Spark come ready to use, with a general-purpose configuration and upgrade management. The study characterizes both the BigBench queries and the out-of-the-box performance of Spark and Hive versions in the cloud. At the same time, comparing popular PaaS offerings in terms of reliability, data scalability (1 GB to 10 TB), versions, and settings from Azure HDinsight, Amazon Web Services EMR, and Google Cloud Dataproc. The query characterization highlights the similarities and differences in Hive an Spark frameworks, and which queries are the most resource consuming according to CPU, memory, and I/O. Scalability results show how there is a need for configuration tuning in most cloud providers as data scale grows, especially with Sparks memory usage. These results can help practitioners to quickly test systems by picking a subset of the queries which stresses each of the categories. At the same time, results show how Hive and Spark compare and what performance can be expected of each in PaaS., This project has received funding from the European Research Council (ERC) under the European Unions Horizon 2020 research and innovation programme (grant agreement No. 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Real-time data center's telemetry reduction and reconstruction using Markov chain models

  • Baig, Shuja-ur-Rehman
  • Iqbal, Waheed
  • Berral García, Josep Lluís
  • Erradi, Abdelkarim
  • Carrera Pérez, David
Large-scale data centers are composed of thousands of servers organized in interconnected racks to offer services to users. These data centers continuously generate large amounts of telemetry data streams (e.g., hardware utilization metrics) used for multiple purposes, including resource management, workload characterization, resource utilization prediction, capacity planning, and real-time analytics. These telemetry streams require costly bandwidth utilization and storage space, particularly at medium-long term for large data centers. This paper addresses this problem by proposing and evaluating a system to efficiently reduce bandwidth and storage for telemetry data through real-time modeling using Markov chain based methods. Our proposed solution was evaluated using real telemetry datasets and compared with polynomial regression methods for reducing and reconstructing data. Experimental results show that data can be lossy compressed up to 75% for bandwidth utilization and 95.33% for storage space, with reconstruction accuracy close to 92%., This work was supported in part by the European Research Council (ERC) under the EU Horizon 2020 programme under Grant GA 639595, in part by the Spanish Ministry of Economy, Industry and Competitiveness under Grant TIN2015-65316-P and Grant IJCI2016-27485, in part by the Generalitat de Catalunya under Grant 2014-SGR-1051, in part by the University of the Punjab, Pakistan, and in part by the Qatar National Research Fund (a member of Qatar Foundation) under NPRP Grant # NPRP9-224-1-049., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Constant-time sliding window framework with reduced memory footprint and efficient bulk evictions

  • Villalba Navarro, Álvaro
  • Berral García, Josep Lluís
  • Carrera Pérez, David
The fast evolution of data analytics platforms has resulted in an increasing demand for real-time data stream processing. From Internet of Things applications to the monitoring of telemetry generated in large data centers, a common demand for currently emerging scenarios is the need to process vast amounts of data with low latencies, generally performing the analysis process as close to the data source as possible. Stream processing platforms are required to be malleable and absorb spikes generated by fluctuations of data generation rates. Data is usually produced as time series that have to be aggregated using multiple operators, being sliding windows one of the most common abstractions used to process data in real-time. To satisfy the above-mentioned demands, efficient stream processing techniques that aggregate data with minimal computational cost need to be developed. In this paper we present the Monoid Tree Aggregator general sliding window aggregation framework, which seamlessly combines the following features: amortized O(1) time complexity and a worst-case of O(log n) between insertions; it provides both a window aggregation mechanism and a window slide policy that are user programmable; the enforcement of the window sliding policy exhibits amortized O(1) computational cost for single evictions and supports bulk evictions with cost O(log n) ; and it requires a local memory space of O(log n) . The framework can compute aggregations over multiple data dimensions, and has been designed to support decoupling computation and data storage through the use of distributed Key-Value Stores to keep window elements and partial aggregations., This project is partially supported by the European Research Council (ERC), Spain under the European Unions Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015- 65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Topology-aware GPU scheduling for learning workloads in cloud environments

  • Amaral, Marcelo
  • Polo Bardés, Jordà
  • Carrera, David
  • Seelam, Seetharami
  • Steinder, Malgorzata
Recent advances in hardware, such as systems with multiple GPUs and their availability in the cloud, are enabling deep learning in various domains including health care, autonomous vehicles, and Internet of Things. Multi-GPU systems exhibit complex connectivity among GPUs and between GPUs and CPUs. Workload schedulers must consider hardware topology and workload communication requirements in order to allocate CPU and GPU resources for optimal execution time and improved utilization in shared cloud environments. This paper presents a new topology-aware workload placement strategy to schedule deep learning jobs on multi-GPU systems. The placement strategy is evaluated with a prototype on a Power8 machine with Tesla P100 cards, showing speedups of up to ≈1.30x compared to state-of-the-art strategies; the proposed algorithm achieves this result by allocating GPUs that satisfy workload requirements while preventing interference. Additionally, a large-scale simulation shows that the proposed strategy provides higher resource utilization and performance in cloud systems., This project is supported by the IBM/BSC Technology Center for Supercomputing collaboration agreement. It has also received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). It is also partially supported by the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya under contract 2014SGR1051, by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493). We thank our IBM Research colleagues Alaa Youssef and Asser Tantawi for the valuable discussions. We also thank SC17 committee member Blair Bethwaite of Monash University for his constructive feedback on the earlier drafts of this paper., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Performance characterization of spark workloads on shared NUMA Systems

  • Baig, Shuja-ur-Rehman
  • Amaral, Marcelo
  • Polo Cantero, José
  • Carrera Pérez, David
As the adoption of Big Data technologies becomes the norm in an increasing number of scenarios, there is also a growing need to optimize them for modern processors. Spark has gained momentum over the last few years among companies looking for high performance solutions that can scale out across different cluster sizes. At the same time, modern processors can be connected to large amounts of physical memory, in the range of up to few terabytes. This opens an enormous range of opportunities for runtimes and applications that aim to improve their performance by leveraging low latencies and high bandwidth provided by RAM. The result is that there are several examples today of applications that have started pushing the in-memory computing paradigm to accelerate tasks. To deliver such a large physical memory capacity, hardware vendors have leveraged Non-Uniform Memory Architectures (NUMA). This paper explores how Spark-based workloads are impacted by the effects of NUMA-placement decisions, how different Spark configurations result in changes in delivered performance, how the characteristics of the applications can be used to predict workload collocation conflicts, and how to improve performance by collocating workloads in scale-up nodes. We explore several workloads run on top of the IBM Power8 processor, and provide manual strategies that can leverage performance improvements up to 40% on Spark workloads when using smart processor-pinning and workload collocation strategies., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Performance Characterization of Spark Workloads on Shared NUMA Systems

  • Baig, Shuja-ur-Rehman
  • Amaral, Marcelo
  • Polo, Jordà
  • Carrera, David
As the adoption of Big Data technologies becomes the norm in an increasing number of scenarios, there is also a growing need to optimize them for modern processors. Spark has gained momentum over the last few years among companies looking for high performance solutions that can scale out across different cluster sizes. At the same time, modern processors can be connected to large amounts of physical memory, in the range of up to few terabytes. This opens an enormous range of opportunities for runtimes and applications that aim to improve their performance by leveraging low latencies and high bandwidth provided by RAM. The result is that there are several examples today of applications that have started pushing the in-memory computing paradigm to accelerate tasks. To deliver such a large physical memory capacity, hardware vendors have leveraged Non-Uniform Memory Architectures (NUMA). This paper explores how Spark-based workloads are impacted by the effects of NUMA-placement decisions, how different Spark configurations result in changes in delivered performance, how the characteristics of the applications can be used to predict workload collocation conflicts, and how to improve performance by collocating workloads in scale-up nodes. We explore several workloads run on top of the IBM Power8 processor, and provide manual strategies that can leverage performance improvements up to 40% on Spark workloads when using smart processor-pinning and workload collocation strategies., This work is partially supported by the European Research Council (ERC) under the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitiveness (TIN2015-65316-P) and the Generalitat de Catalunya (2014-SGR-1051)., Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Constant-Time Sliding Window Framework with Reduced Memory Footprint and Efficient Bulk Evictions

  • Villalba, Álvaro
  • Berral García, Josep Lluís
  • Carrera, David
The fast evolution of data analytics platforms has resulted in an increasing demand for real-time data stream processing. From Internet of Things applications to the monitoring of telemetry generated in large data centers, a common demand for currently emerging scenarios is the need to process vast amounts of data with low latencies, generally performing the analysis process as close to the data source as possible. Stream processing platforms are required to be malleable and absorb spikes generated by fluctuations of data generation rates. Data is usually produced as time series that have to be aggregated using multiple operators, being sliding windows one of the most common abstractions used to process data in real-time. To satisfy the above-mentioned demands, efficient stream processing techniques that aggregate data with minimal computational cost need to be developed. In this paper we present the Monoid Tree Aggregator general sliding window aggregation framework, which seamlessly combines the following features: amortized O(1) time complexity and a worst-case of O(logn) between insertions; it provides both a window aggregation mechanism and a window slide policy that are user programmable; the enforcement of the window sliding policy exhibits amortized O(1) computational cost for single evictions and supports bulk evictions with cost O(logn) ; and it requires a local memory space of O(logn) . The framework can compute aggregations over multiple data dimensions, and has been designed to support decoupling computation and data storage through the use of distributed Key-Value Stores to keep window elements and partial aggregations., This project is partially supported by the European ResearchCouncil (ERC), Spain under the European Unions Horizon2020 research and innovation programme (grant agreementNo 639595). It is also partially supported by the Ministryof Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051,by the ICREA Academia program, and by the BSC-CNS Severo Ochoa program (SEV-2015-0493)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Adaptive sliding windows for improved estimation of data center resource utilization

  • Baig, Shuja-ur-Rehman
  • Iqbal, Waheed
  • Berral García, Josep Lluís
  • Carrera Pérez, David
Accurate prediction of data center resource utilization is required for capacity planning, job scheduling, energy saving, workload placement, and load balancing to utilize the resources efficiently. However, accurately predicting those resources is challenging due to dynamic workloads, heterogeneous infrastructures, and multi-tenant co-hosted applications. Existing prediction methods use fixed size observation windows which cannot produce accurate results because of not being adaptively adjusted to capture local trends in the most recent data. Therefore, those methods train on large fixed sliding windows using an irrelevant large number of observations yielding to inaccurate estimations or fall for inaccuracy due to degradation of estimations with short windows on quick changing trends. In this paper we propose a deep learning-based adaptive window size selection method, dynamically limiting the sliding window size to capture the trend for the latest resource utilization, then build an estimation model for each trend period. We evaluate the proposed method against multiple baseline and state-of-the-art methods, using real data-center workload data sets. The experimental evaluation shows that the proposed solution outperforms those state-of-the-art approaches and yields 16 to 54% improved prediction accuracy compared to the baseline methods., This work is partially supported by the European ResearchCouncil (ERC) under the EU Horizon 2020 programme(GA 639595), the Spanish Ministry of Economy, Industry andCompetitiveness (TIN2015-65316-P and IJCI2016-27485), theGeneralitat de Catalunya, Spain (2014-SGR-1051) and Universityof the Punjab, Pakistan. The statements made herein are solelythe responsibility of the authors., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Disaggregating non-volatile memory for throughput-oriented genomics workloads

  • Call Barreiro, Erin
  • Polo Bardés, Jordà
  • Carrera Pérez, David
  • Guim Bernat, Francesc
  • Sen, Sujoy
Massive exploitation of next-generation sequencing technologies requires dealing with both: huge amounts of data and complex bioinformatics pipelines. Computing architectures have evolved to deal with these problems, enabling approaches that were unfeasible years ago: accelerators and Non-Volatile Memories (NVM) are becoming widely used to enhance the most demanding workloads. However, bioinformatics workloads are usually part of bigger pipelines with different and dynamic needs in terms of resources. The introduction of Software Defined Infrastructures (SDI) for data centers provides roots to dramatically increase the efficiency in the management of infrastructures. SDI enables new ways to structure hardware resources through disaggregation, and provides new hardware composability and sharing mechanisms to deploy workloads in more flexible ways. In this paper we study a state-of-the-art genomics application, SMUFIN, aiming to address the challenges of future HPC facilities., This work is partially supported by the European Research Council (ERC) under the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitivity (TIN2015-65316-P) and the Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Performance evaluation of microservices architectures using containers

  • Amaral, Marcelo
  • Polo Bardés, Jordà
  • Carrera Pérez, David
  • Mohomed, Iqbal
  • Unuvar, Merve
  • Steinder, Malgorzata
Microservices architecture has started a new trend for application development for a number of reasons: (1) to reduce complexity by using tiny services; (2) to scale, remove and deploy parts of the system easily; (3) to improve flexibility to use different frameworks and tools; (4) to increase the overall scalability; and (5) to improve the resilience of the system. Containers have empowered the usage of microservices architectures by being lightweight, providing fast start-up times, and having a low overhead. Containers can be used to develop applications based on monolithic architectures where the whole system runs inside a single container or inside a microservices architecture where one or few processes run inside the containers. Two models can be used to implement a microservices architecture using containers: master-slave, or nested-container. The goal of this work is to compare the performance of CPU and network running benchmarks in the two aforementioned models of microservices architecture hence provide a benchmark analysis guidance for system designers., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Improving maritime traffic emission estimations on missing data with CRBMs

  • Gutiérrez Torre, Alberto
  • Berral García, Josep Lluís
  • Buchaca Prats, David
  • Guevara Vilardell, Marc
  • Soret, Albert
  • Carrera Pérez, David
© Elsevier 2020. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/, Maritime traffic emissions are a major concern to governments as they heavily impact the Air Quality in coastal cities. Ships use the Automatic Identification System (AIS) to continuously report position and speed among other features, and therefore this data is suitable to be used to estimate emissions, if it is combined with engine data. However, important ship features are often inaccurate or missing. State-of-the-art complex systems, like CALIOPE at the Barcelona Supercomputing Center, are used to model Air Quality. These systems can benefit from AIS based emission models as they are very precise in positioning the pollution. Unfortunately, these models are sensitive to missing or corrupted data, and therefore they need data curation techniques to significantly improve the estimation accuracy. In this work, we propose a methodology for treating ship data using Conditional Restricted Boltzmann Machines (CRBMs) plus machine learning methods to improve the quality of data passed to emission models that can also be applied to other GPS and time-series problems. Results show that we can improve the default methods proposed to cover missing data. In our results, we observed that using our method the models boosted their accuracy to detect otherwise undetectable emissions. In particular, we used a real data-set of AIS data, provided by the Spanish Port Authority, to estimate that thanks to our method, the model was able to detect 45% of additional emissions, representing 152 tonnes of pollutants per week in Barcelona and propose new features that may enhance emission modeling., This project hasreceived funding from the European Research Council (ERC) under the European Union Horizon 2020 research and innovation programme (grant agreement No 639595). This work is partially supported by the Ministry of Economy, Industry and Competitiveness of Spain under contracts TIN2015-65316-P,2014SGR1051, IJCI2016-27485, and Severo Ochoa Center of Excellence SEV-2015-0493-16-5., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Considerations in using OpenCL on GPUs and FPGAs for throughput-oriented genomics workloads

  • Cadenelli, Nicola
  • Jaksic, Zoran
  • Polo Bardés, Jordà
  • Carrera, David
The recent upsurge in the available amount of health data and the advances in next-generation sequencing are setting the ground for the long-awaited precision medicine. To process this deluge of data, bioinformatics workloads are becoming more complex and more computationally demanding. For this reasons they have been extended to support different computing architectures, such as GPUs and FPGAs, to leverage the form of parallelism typical of each of such architectures. The paper describes how a genomic workload such as k-mer frequency counting that takes advantage of a GPU can be offloaded to one or even more FPGAs. Moreover, it performs a comprehensive analysis of the FPGA acceleration comparing its performance to a non-accelerated configuration and when using a GPU. Lastly, the paper focuses on how, when using accelerators with a throughput-oriented workload, one should also take into consideration both kernel execution time and how well each accelerator board overlaps kernels and PCIe transferred. Results show that acceleration with two FPGAs can improve both time- and energy-to-solution for the entire accelerated part by a factor of 1.32x. Per contra, acceleration with one GPU delivers an improvement of 1.77x in time-to-solution but of a lower 1.49x in energy-to-solution due to persistently higher power consumption. The paper also evaluates how future FPGA boards with components (i.e., off-chip memory and PCIe) on par with those of the GPU board could provide an energy-efficient alternative to GPUs., This work was supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement s No 639595); the Ministry of Economy of Spain under contract TIN2015-65316-P and Generalitat de Catalunya, Spain under contract 2014SGR1051; the ICREA, Spain Academia program; and the BSC-CNS Severo Ochoa, Spain program (SEV-2015-0493)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Multi-tenant Pub/Sub processing for real-time data streams

  • Villalba Navarro, Álvaro
  • Carrera Pérez, David
Devices and sensors generate streams of data across a diversity of locations and protocols. That data usually reaches a central platform that is used to store and process the streams. Processing can be done in real time, with transformations and enrichment happening on-the-fly, but it can also happen after data is stored and organized in repositories. In the former case, stream processing technologies are required to operate on the data; in the latter batch analytics and queries are of common use. This paper introduces a runtime to dynamically construct data stream processing topologies based on user-supplied code. These dynamic topologies are built on-the-fly using a data subscription model defined by the applications that consume data. Each user-defined processing unit is called a Service Object. Every Service Object consumes input data streams and may produce output streams that others can consume. The subscription-based programing model enables multiple users to deploy their own data-processing services. The runtime does the dynamic forwarding of data and execution of Service Objects from different users. Data streams can originate in real-world devices or they can be the outputs of Service Objects. The runtime leverages Apache STORM for parallel data processing, that combined with dynamic user-code injection provides multi-tenant stream processing topologies. In this work we describe the runtime, its features and implementation details, as well as we include a performance evaluation of some of its core components., This work is partially supported by the European Research Council (ERC) un- der the EU Horizon 2020 programme (GA 639595), the Spanish Ministry of Economy, Industry and Competitivity (TIN2015-65316-P) and the Generalitat de Catalunya (2014-SGR-1051)., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

ALOJA: A framework for benchmarking and predictive analytics in Hadoop deployments

  • Berral García, Josep Lluís
  • Poggi Mastrokalo, Nicolas
  • Carrera Pérez, David
  • Call Barreiro, Erin
  • Reinauer, Rob
  • Green, Daron
This article presents the ALOJA project and its analytics tools, which leverages machine learning to interpret Big Data benchmark performance data and tuning. ALOJA is part of a long-term collaboration between BSC and Microsoft to automate the characterization of cost-effectiveness on Big Data deployments, currently focusing on Hadoop. Hadoop presents a complex run-time environment, where costs and performance depend on a large number of configuration choices. The ALOJA project has created an open, vendor-neutral repository, featuring over 40,000 Hadoop job executions and their performance details. The repository is accompanied by a test-bed and tools to deploy and evaluate the cost-effectiveness of different hardware configurations, parameters and Cloud services. Despite early success within ALOJA, a comprehensive study requires automation of modeling procedures to allow an analysis of large and resource-constrained search spaces. The predictive analytics extension, ALOJA-ML, provides an automated system allowing knowledge discovery by modeling environments from observed executions. The resulting models can forecast execution behaviors, predicting execution times for new configurations and hardware choices. That also enables model-based anomaly detection or efficient benchmark guidance by prioritizing executions. In addition, the community can benefit from ALOJA data-sets and framework to improve the design and deployment of Big Data applications., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 639595). This work is partially supported by the Ministry of Economy of Spain under contracts TIN2012-34557 and 2014SGR1051., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Considerations in using OpenCL on GPUs and FPGAs for throughput-oriented genomics workloads

  • Cadenelli, Nicola
  • Jaksic, Zoran
  • Polo Bardés, Jordà
  • Carrera Pérez, David
The recent upsurge in the available amount of health data and the advances in next-generation sequencing are setting the ground for the long-awaited precision medicine. To process this deluge of data, bioinformatics workloads are becoming more complex and more computationally demanding. For this reasons they have been extended to support different computing architectures, such as GPUs and FPGAs, to leverage the form of parallelism typical of each of such architectures. The paper describes how a genomic workload such as k-mer frequency counting that takes advantage of a GPU can be offloaded to one or even more FPGAs. Moreover, it performs a comprehensive analysis of the FPGA acceleration comparing its performance to a non-accelerated configuration and when using a GPU. Lastly, the paper focuses on how, when using accelerators with a throughput-oriented workload, one should also take into consideration both kernel execution time and how well each accelerator board overlaps kernels and PCIe transferred. Results show that acceleration with two FPGAs can improve both time- and energy-to-solution for the entire accelerated part by a factor of 1.32x. Per contra, acceleration with one GPU delivers an improvement of 1.77x in time-to-solution but of a lower 1.49x in energy-to-solution due to persistently higher power consumption. The paper also evaluates how future FPGA boards with components (i.e., off-chip memory and PCIe) on par with those of the GPU board could provide an energy-efficient alternative to GPUs., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Sequence-to-sequence models for workload interference prediction on batch processing datacenters

  • Buchaca Prats, David
  • Marcual Medina, Joan
  • Berral García, Josep Lluís
  • Carrera Pérez, David
Co-scheduling of jobs in data centers is a challenging scenario where jobs can compete for resources, leading to severe slowdowns or failed executions. Efficient job placement on environments where resources are shared requires awareness on how jobs interfere during execution, to go far beyond ineffective resource overbooking techniques. Current techniques, most of which already involve machine learning and job modeling, are based on workload behavior summarization over time, rather than focusing on effective job requirements at each instant of the execution. In this work, we propose a methodology for modeling co-scheduling of jobs on data centers, based on their behavior towards resources and execution time and using sequence-to-sequence models based on recurrent neural networks. The goal is to forecast co-executed jobs footprint on resources throughout their execution time, from the profile shown by the individual jobs, in order to enhance resource manager and scheduler placement decisions. The methods presented herein are validated by using High Performance Computing benchmarks based on different frameworks (such as Hadoop and Spark) and applications (CPU bound, IO bound, machine learning, SQL queries...). Experiments show that the model can correctly identify the resource usage trends from previously seen and even unseen co-scheduled jobs., This work is supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement no. 639595); [Generalitat] de Catalunya under contract 2014SGR1051; the ICREA Academia program; and the BSC-CNS Severo Ochoa program (SEV-2015-0493); the Spanish Ministry of Economy under contract TIN2015-65316-P and the Generalitat., Peer Reviewed, Postprint (author's final draft)