Benjamin Gaidioz

Managing ATLAS data on a petabyte-scale with DQ2

The ATLAS detector at CERNs Large Hadron Collider presents data handling requirements on an unprecedented scale. From 2008 on the ATLAS distributed data management system, Don Quijote2 (DQ2), must manage tens of petabytes of experiment data per year, distributed globally via the LCG, OSG and NDGF computing grids, now commonly known as the WLCG. Since its inception in 2005 DQ2 has continuously managed all experiment data for the ATLAS collaboration, which now comprises over 3000 scientists participating from more than 150 universities and laboratories in 34 countries. Fulfilling its primary requirement of providing a highly distributed, fault-tolerant and scalable architecture DQ2 was successfully upgraded from managing data on a terabyte-scale to managing data on a petabyte-scale. We present improvements and enhancements to DQ2 based on the increasing demands for ATLAS data management. We describe performance issues, architectural changes and implementation decisions, the current state of deployment in test and production as well as anticipated future improvements. Test results presented here show that DQ2 is capable of handling data up to and beyond the requirements of full-scale data-taking.

Exploring high performance distributed file storage using LDPC codes

We explore the feasibility of implementing a reliable, high performance, distributed storage system on a commodity computing cluster. Files are distributed across storage nodes using erasure coding with small low-density parity-check (LDPC) codes, which provide high-reliability with small storage and performance overhead. We present performance measurements done on a prototype system comprising 50 nodes, which are self organised using a peer-to-peer overlay.

Experiment Dashboard: the monitoring system for the LHC experiments

In this paper we present the Experiment Dashboard monitoring system, which is currently in use by four Large Hadron Collider (LHC)[1] experiments. The goal of the Experiment Dashboard is to monitor the activities of the LHC experiments on the distributed infrastructure, providing monitoring data from the virtual organization/user perspectives. All LHC experiments are using various Grid infrastructures (LCG/EGEE, OSG, Nordugrid) with correspondingly various middleware flavors and job submission methods. Providing a uniform and complete view of the various activities like job processing, data movement and publishing, access to distributed databases regardless of the underlying Grid flavor is the challenging task. In this paper we will describe the solutions implemented in the Experiment Dashboard.

CMS Dashboard Task Monitoring: A user-centric monitoring view

We are now in a phase change of the CMS experiment where people are turning more intensely to physics analysis and away from construction. This brings a lot of challenging issues with respect to monitoring of the user analysis. The physicists must be able to monitor the execution status, application and grid-level messages of their tasks that may run at any site within the CMS Virtual Organisation. The CMS Dashboard Task Monitoring project provides this information towards individual analysis users by collecting and exposing a user-centric set of information regarding submitted tasks including reason of failure, distribution by site and over time, consumed time and efficiency. The development was user-driven with physicists invited to test the prototype in order to assemble further requirements and identify weaknesses with the application.

Experiment Dashboard for Monitoring Computing Activities of the LHC Virtual Organizations

The Large Hadron Collider (LHC) is preparing for data taking at the end of 2009. The Worldwide LHC Computing Grid (WLCG) provides data storage and computational resources for the high energy physics community. Operating the heterogeneous WLCG infrastructure, which integrates 140 computing centers in 33 countries all over the world, is a complicated task. Reliable monitoring is one of the crucial components of the WLCG for providing the functionality and performance that is required by the LHC experiments. The Experiment Dashboard system provides monitoring of the WLCG infrastructure from the perspective of the LHC experiments and covers the complete range of their computing activities. This work describes the architecture of the Experiment Dashboard system and its main monitoring applications and summarizes current experiences by the LHC experiments, in particular during service challenges performed on the WLCG over the last years.

Distributed analysis in ATLAS using GANGA

Distributed data analysis using Grid resources is one of the fundamental applications in high energy physics to be addressed and realized before the start of LHC data taking. The needs to manage the resources are very high. In every experiment up to a thousand physicists will be submitting analysis jobs to the Grid. Appropriate user interfaces and helper applications have to be made available to assure that all users can use the Grid without expertise in Grid technology. These tools enlarge the number of Grid users from a few production administrators to potentially all participating physicists. The GANGA job management system (http://cern.ch/ganga), developed as a common project between the ATLAS and LHCb experiments, provides and integrates these kind of tools. GANGA provides a simple and consistent way of preparing, organizing and executing analysis tasks within the experiment analysis framework, implemented through a plug-in system. It allows trivial switching between running test jobs on a local batch system and running large-scale analyzes on the Grid, hiding Grid technicalities. We will be reporting on the plug-ins and our experiences of distributed data analysis using GANGA within the ATLAS experiment. Support for all Grids presently used by ATLAS, namely the LCG/EGEE, NDGF/NorduGrid, and OSG/PanDA is provided. The integration and interaction with the ATLAS data management system DQ2 into GANGA is a key functionality. An intelligent job brokering is set up by using the job splitting mechanism together with data-set and file location knowledge. The brokering is aided by an automated system that regularly processes test analysis jobs at all ATLAS DQ2 supported sites. Large numbers of analysis jobs can be sent to the locations of data following the ATLAS computing model. GANGA supports amongst other things tasks of user analysis with reconstructed data and small scale production of Monte Carlo data.

Job monitoring on the WLCG scope: Current status and new strategy

Job processing and data transfer are the main computing activities on the WLCG infrastructure. Reliable monitoring of the job processing on the WLCG scope is a complicated task due to the complexity of the infrastructure itself and the diversity of the currently used job submission methods. The paper will describe current status and the new strategy for the job monitoring on the WLCG scope, covering primary information sources, job status changes publishing, transport mechanism and visualization.

Dashboard applications to monitor experiment activities at sites

In the framework of a distributed computing environment, such as WLCG, monitoring has a key role in order to keep under control activities going on in sites located in different countries and involving people based in many different sites. To be able to cope with such a large scale heterogeneous infrastructure, it is necessary to have monitoring tools providing a complete and reliable view of the overall performance of the sites. Moreover, the structure of a monitoring system critically depends on the object to monitor and on the users it is addressed to. In this article we will describe two different monitoring systems both aimed to monitor activities and services provided in the WLCG framework, but designed in order to meet the requirements of different users: Site Status Board has an overall view of the services available in all the sites supporting an experiment, whereas Siteview provides a complete view of all the activities going on at a site, for all the experiments supported by the site.

High performance event-building in linux for LHCb

The LHCb experiment event-building is performed over a Gigabit Ethernet switched network. One specific step of event-building is implemented by a software running on a gateway PC whose role is to gather data packets from data sources, rebuild events and forward them to computing nodes for running trigger algorithms. In this article, we concentrate on the implementation of this component on a Linux system. While implementing the software, we made thorough studies of the kernel and profiled applications, leading to significant performance improvement. More importantly, these studies allowed us to also gain in terms of predictability thanks to a good understanding of the whole system. In this article, we use this application to illustrate possible improvements to system software for data acquisition. We describe in detail implementation choices and related operating system kernel code. These techniques and observations are generic enough to be applied to other similar systems.

Optimization of event-building implementation on top of gigabit Ethernet

LHCb is implementing event-building over a gigabit Ethernet switched network. One component of the data acquisition network is a gateway PC which role is to gather data packets from data sources, rebuild events and forward them to computing nodes for trigger algorithms. In this article, we concentrate on the implementation of this component as a software running on a Linux system. We describe in details implementation choices and details of the operating system kernel which permitted us to achieve a higher performance and ensure a higher predictability of the system