CERN - IT Department CH-1211 Genève 23 Switzerland www.cern.ch/i t The High Performance Archiver for the LHC Experiments Manuel Gonzalez Berges CERN, Geneva.

Slides:



Advertisements
Similar presentations
JCOP FW Update ALICE DCS Workshop 6 th and 7 th October, 2005 Fernando Varela Rodriguez, IT-CO Outline Organization Current status Future work.
Advertisements

PVSS and JCOP Framework Organization, Support & News Oliver Holme IT-CO.
The Control System for the ATLAS Pixel Detector
Overview of Database Administrator (DBA) Tools
Oracle9i Database Administrator: Implementation and Administration 1 Chapter 2 Overview of Database Administrator (DBA) Tools.
Detector Control System at the GIF++ A. Polini, M. Romano (AIDA+INFN Bologna) on behalf of WP group Gif++ User Meeting 11/07/2014 Outline Status.
CERN IT Department CH-1211 Geneva 23 Switzerland t Marcin Blaszczyk, IT-DB Atlas standby database tests February.
André Augustinus ALICE Detector Control System  ALICE DCS is responsible for safe, stable and efficient operation of the experiment  Central monitoring.
CHEP 2012 – New York City 1.  LHC Delivers bunch crossing at 40MHz  LHCb reduces the rate with a two level trigger system: ◦ First Level (L0) – Hardware.
1 Databases in ALICE L.Betev LCG Database Deployment and Persistency Workshop Geneva, October 17, 2005.
23/04/2008VLVnT08, Toulon, FR, April 2008, M. Stavrianakou, NESTOR-NOA 1 First thoughts for KM3Net on-shore data storage and distribution Facilities VLV.
L. Granado Cardoso, F. Varela, N. Neufeld, C. Gaspar, C. Haen, CERN, Geneva, Switzerland D. Galli, INFN, Bologna, Italy ICALEPCS, October 2011.
CERN - IT Department CH-1211 Genève 23 Switzerland t Partitioning in COOL Andrea Valassi (CERN IT-DM) R. Basset (CERN IT-DM) Distributed.
Scale-out databases for CERN use cases Strata Hadoop World London 6 th of May,2015 Zbigniew Baranowski, CERN IT-DB.
Database Services for Physics at CERN with Oracle 10g RAC HEPiX - April 4th 2006, Rome Luca Canali, CERN.
CERN - IT Department CH-1211 Genève 23 Switzerland t Monitoring the ATLAS Distributed Data Management System Ricardo Rocha (CERN) on behalf.
Scaling Up PVSS Phase II. 2 Purpose of this talk Start a discussion about the next phase of the Scaling Up PVSS Project. Start a discussion about the.
Performance Concepts Mark A. Magumba. Introduction Research done on 1058 correspondents in 2006 found that 75% OF them would not return to a website that.
CERN IT Department CH-1211 Geneva 23 Switzerland t The Experiment Dashboard ISGC th April 2008 Pablo Saiz, Julia Andreeva, Benjamin.
09/11/20061 Detector Control Systems A software implementation: Cern Framework + PVSS Niccolo’ Moggi and Stefano Zucchelli University and INFN Bologna.
CERN - IT Department CH-1211 Genève 23 Switzerland t Tier0 database extensions and multi-core/64 bit studies Maria Girone, CERN IT-PSS LCG.
Clara Gaspar, October 2011 The LHCb Experiment Control System: On the path to full automation.
Update on Database Issues Peter Chochula DCS Workshop, June 21, 2004 Colmar.
Jean-Yves Nief CC-IN2P3, Lyon HEPiX-HEPNT, Fermilab October 22nd – 25th, 2002.
Experiment Support CERN IT Department CH-1211 Geneva 23 Switzerland t DBES PhEDEx Monitoring Nicolò Magini CERN IT-ES-VOS For the PhEDEx.
The Joint COntrols Project Framework Manuel Gonzalez Berges on behalf of the JCOP FW Team.
Databases E. Leonardi, P. Valente. Conditions DB Conditions=Dynamic parameters non-event time-varying Conditions database (CondDB) General definition:
6/26/01High Throughput Linux Clustering at Fermilab--S. Timm 1 High Throughput Linux Clustering at Fermilab Steven C. Timm--Fermilab.
Achieving Scalability, Performance and Availability on Linux with Oracle 9iR2-RAC Grant McAlister Senior Database Engineer Amazon.com Paper
André Augustinus 10 October 2005 ALICE Detector Control Status Report A. Augustinus, P. Chochula, G. De Cataldo, L. Jirdén, S. Popescu the DCS team, ALICE.
ALICE, ATLAS, CMS & LHCb joint workshop on
Management of the LHCb DAQ Network Guoming Liu * †, Niko Neufeld * * CERN, Switzerland † University of Ferrara, Italy.
6 May 2014 CERN openlab IT Challenges workshop, Kacper Szkudlarek, CERN Manuel.
CERN - IT Department CH-1211 Genève 23 Switzerland t Oracle Real Application Clusters (RAC) Techniques for implementing & running robust.
CERN - IT Department CH-1211 Genève 23 Switzerland t COOL Conditions Database for the LHC Experiments Development and Deployment Status Andrea.
Peter Chochula ALICE Offline Week, October 04,2005 External access to the ALICE DCS archives.
Mark E. Fuller Senior Principal Instructor Oracle University Oracle Corporation.
JCOP/DCS Plans for HW Upgrade A. Augustinus (ALICE), C
EGEE-II INFSO-RI Enabling Grids for E-sciencE EGEE and gLite are registered trademarks Implementation and performance analysis of.
CERN Database Services for the LHC Computing Grid Maria Girone, CERN.
Copyright 2007, Information Builders. Slide 1 Machine Sizing and Scalability Mark Nesson, Vashti Ragoonath June 2008.
CERN IT Department CH-1211 Genève 23 Switzerland t DBA Experience in a multiple RAC environment DM Technical Meeting, Feb 2008 Miguel Anjo.
CERN - IT Department CH-1211 Genève 23 Switzerland t High Availability Databases based on Oracle 10g RAC on Linux WLCG Tier2 Tutorials, CERN,
Database Competence Centre openlab Major Review Meeting nd February 2012 Maaike Limper Zbigniew Baranowski Luigi Gallerani Mariusz Piorkowski Anton.
Management of the LHCb Online Network Based on SCADA System Guoming Liu * †, Niko Neufeld † * University of Ferrara, Italy † CERN, Geneva, Switzerland.
Alarm Handling Oliver Holme 7 th November Guidelines & Strategies All provide recommendation to use Framework Classes Standard severities/behaviour/colours.
Distributed Logging Facility Castor External Operation Workshop, CERN, November 14th 2006 Dennis Waldron CERN / IT.
Scalable data access with Impala Zbigniew Baranowski Maciej Grzybek Daniel Lanza Garcia Kacper Surdy.
Clara Gaspar, April 2006 LHCb Experiment Control System Scope, Status & Worries.
LHCb Configuration Database Lana Abadie, PhD student (CERN & University of Pierre et Marie Curie (Paris VI), LIP6.
Management of the LHCb DAQ Network Guoming Liu *†, Niko Neufeld * * CERN, Switzerland † University of Ferrara, Italy.
The DCS Databases Peter Chochula. 31/05/2005Peter Chochula 2 Outline PVSS basics (boring topic but useful if one wants to understand the DCS data flow)
RAC aware change In order to reduce the cluster contention, a new scheme for the insertion has been developed. In the new scheme: - each “client” receives.
Markus Frank (CERN) & Albert Puig (UB).  An opportunity (Motivation)  Adopted approach  Implementation specifics  Status  Conclusions 2.
Database Issues Peter Chochula 7 th DCS Workshop, June 16, 2003.
© 2006 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice Database Growth: Problems & Solutions.
Considerations for database servers Castor review – June 2006 Eric Grancher, Nilo Segura Chinchilla IT-DES.
JCOP Framework and PVSS News ALICE DCS Workshop 14 th March, 2006 Piotr Golonka CERN IT/CO-BE Outline PVSS status Framework: Current status and future.
The ALICE data quality monitoring Barthélémy von Haller CERN PH/AID For the ALICE Collaboration.
M. Caprini IFIN-HH Bucharest DAQ Control and Monitoring - A Software Component Model.
CERN - IT Department CH-1211 Genève 23 Switzerland t ASM and Oracle Service Availability Monitoring LCG 3D Workshop CERN, January 26 th,
CERN IT Department CH-1211 Genève 23 Switzerland t Load testing & benchmarks on Oracle RAC Romain Basset – IT PSS DP.
1 PVSS Oracle scalability Target = changes per second (tested with 160k) changes per client 5 nodes RAC NAS 3040, each with one.
The PVSS Oracle Archiver FW WG 6 th July Credits Many people involved IT/DES: Eric Grancher, Nilo Segura, Chris Lambert IT/PSS: Luca Canali ALICE:
PVSS an industrial tool for slow control
DCS Status and Amanda News
PVSS Evolution in Relation to Databases
Oracle Storage Performance Studies
Case studies – Atlas and PVSS Oracle archiver
Support for ”interactive batch”
Presentation transcript:

CERN - IT Department CH-1211 Genève 23 Switzerland t The High Performance Archiver for the LHC Experiments Manuel Gonzalez Berges CERN, Geneva (Switzerland)

CERN IT Department CH-1211 Genève 23 Switzerland t Controls The High Performance Archiver for the LHC Experiments - 2 Outline Context Archiving in PVSS PVSS Client Database Server Conclusions

Controls The High Performance Archiver for the LHC Experiments - 3 The LHC Experiments Worldwide collaborations One order of magnitude bigger than previous generation (~1 million i/o) Complex operation –Many subdetectors and subsystems Lifetime of 20 years Common approach to controls (Joint Controls Project – JCOP) Currently installing and commissioning ALICE ATLAS CMS LHCb

Controls The High Performance Archiver for the LHC Experiments - 4 LHC experiment control system Back End –Linux/Windows PCs (~150) –PVSS + JCOP Framework –Main functions Graphical user interfaces Operation as a set of Finite State Machines (FSM) Alarm handling Interface to external systems Archiving Front End –Several technologies PCs, PLCs, Embbeded computers, etc –Main functions Data acquisition Filtering Real time loops & FSM Interlocks Detector hardware Control System Front End Back End

Controls The High Performance Archiver for the LHC Experiments - 5 LHC experiments archiving Purpose –Debugging of the control system –Operation –Physics offline analysis (conditions data) Requirements –Data storage Single computer peak rate of 2000 changes/s Full application sustained rate of ~ changes/s (~150 computers at 1000 changes/s) –Data retrieval Optimize known common queries –Get a set of values for a specified time range –Snapshot at a given time Other queries will come with usage

Controls The High Performance Archiver for the LHC Experiments - 6 Archiving in PVSS PVSS system Front End/External systems Monitored values Commands/ Settings Alarms Storage Commands/ Settings

Controls The High Performance Archiver for the LHC Experiments - 7 File vs Database archiver in PVSS File archiver –Local hardisk of each computer –Issues with managing many big files –Proprietary format Database archiver –Centralized server –Relational database –Initially developed for some specific customers Performance far from the extreme requirements of the LHC experiments Close collaboration ETM – CERN to improve it –CERN expertise in databases –Facilities to test on very large systems

Controls The High Performance Archiver for the LHC Experiments - 8 Tables

Controls The High Performance Archiver for the LHC Experiments - 9 Single System Archiving (I) … Event Manager Data Manager Archive Manager PVSS System User Interface Control Manager Driver Oracle Database Server Simulation file …

Controls The High Performance Archiver for the LHC Experiments - 10 Single System Archiving (II) Test results –~ changes/s depending on setup –Bottleneck in Archive Manager –Generic interface to database –Value changes sent one by one Improvements –Use of Oracle native libraries (OCCI) –Bulk insertion Client sends blocks to DB A function inside the DB inserts the data into the history tables –Reduce number of connections to DB One permanent for storage/retrieval One temporary for configuration –Possible to store 2000 changes/s continuously PVSS System Event Manager Data Manager Archive Manager User Interface Driver …

Controls The High Performance Archiver for the LHC Experiments - 11 Test load on the database server PVSS System n PVSS System 2 Event Manager Data Manager Archive Manager PVSS System 1 User Interface Oracle Database Server Driver … Simulation file …

Controls The High Performance Archiver for the LHC Experiments - 12 Performance of Oracle server Oracle server hardware –PC with two Xeon processors, 3 GHz, 4 GBytes RAM, Red Hat Enterprise 4 Tests results –Server can handle ~20-30 systems each at 1000 changes/s –High CPU load Options for improvement –Server with better hardware High cost, limited improvement –Clustered server Low Cost (PCs) Redundancy -> high availability Easy to upgrade by adding more nodes Issues with scalability

Controls The High Performance Archiver for the LHC Experiments - 13 Test setup with RAC server PVSS System 150 PVSS System 2 … Event Manager Data Manager Archive Manager PVSS System 1 User Interface Driver … Oracle Real Application Cluster (RAC) Server Instance n … Instance 2 Database Instance 1 Simulation file …

Controls The High Performance Archiver for the LHC Experiments - 14 Test results with RAC Initial test results –4 server nodes could handle ~ 50 systems –Does not scale linearly The nodes interfere with each other Main improvements –Direct path insertion Constraints disabled during insertion Can be inefficient in space usage –History table partitioned per PVSS client Isolates clients Final performance –Scalable server 6 nodes could handle 150 clients at 1000 changes/s each Possible to add more nodes –Issue with allocation of new storage space (tablespace creation)

Controls The High Performance Archiver for the LHC Experiments - 15 Outlook Values –Insertion Prototype for issue with new space allocation –Space preallocated in advance or with background job –Queries Continue with optimization of known queries Develop an API for external programs Alarms –Bottlenecks identified –ETM working on them –Not extreme requirements Buffering to disk when DB not reachable

Controls The High Performance Archiver for the LHC Experiments - 16 Conclusions LHC experiments archiving requirements are met –Improvements included in standard PVSS version since earlier this year –Issues with retrieval and alarms understood New version expected for the end of the year Fruitful collaboration ETM-CERN –Benefit from CERN expertise on databases –Many solutions were explored with real size test bench –Maintenance guaranteed by the company The archiver is currently in use for the commissioning of the LHC experiments

Controls The High Performance Archiver for the LHC Experiments - 17 Acknowledgements –CERN Chris Lambert Eric Grancher Laura Fernandez Luca Canali Milosz Hulboj Nilo Segura Pior Golonka Svetozar Kapusta –ETM Ewald Sperrer Ronald Putz –Oracle Lothar Flatz The following people have been involved in the work presented:

Controls The High Performance Archiver for the LHC Experiments - 18 Questions?