13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data...

26

Transcript of 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data...

Page 1: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.
Page 2: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

Information discovery in research (CERN)

13 October 2014

Eric Grancher, head of database services, CERN ITManuel Martin Marquez, data scientist, CERN openlab

Page 3: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

3

Outlook• CERN • History of using Oracle and current usage• Information Discovery (Manuel)

Page 4: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

A European Laboratory with Global reach

4

Research

Technology

Science dissemination

Scientific collaborations

Page 5: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

Events at LHC

J. Knobloch/CERN: European Grid Initiative – EGI

Luminosity :1034cm-2 s-1

40 MHz – every 25 ns

20 events overlaying

Page 6: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

Trigger & Data Acquisition

J. Knobloch/CERN: European Grid Initiative – EGI

Page 7: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

Data Recording

J. Knobloch/CERN: European Grid Initiative – EGI

Page 8: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

8

Computing and storage needs• Data volume

• 25 PB per year (in files)• > 5.3 * 1012 rows in an Oracle table (IOT, in one of the databases)

• Computing and storage capacity, world-wide distributed• > 150 sites (grid computing)• > 260 000 CPU cores• > 269 PB disk capacity• > 210 PB tape capacity

• Distributed analysis with costs spread in the different sites (« LHC Computing Grid »)

Page 9: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

9

Oracle at CERN, 1982 accelerator controlhttp://cds.cern.ch/record/443114?ln=en

now 11.2.0.4 and 12.1.0.1

Page 10: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

10

Credit: C. Roderick

Accelerator applications

Page 11: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

11

Accelerator logging50TB/year, rate to increase to 100 – 150 TB in 2014

(Quench ProtectionSystem)

Credit: C. Roderick

Page 12: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

12

Administrative systems

• AIS has standardized on Oracle as database and uses the Oracle databaseas interface between the tools

• Java EE and Apex, deployment with Weblogic• Oracle E-Business HR, managed as Oracle

Managed Cloud Services (“OnDemand”)

Page 13: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

Engineering applications• An integrated PLM platform based on commercial tools.• Simplified web interfaces for precise tasks

3D CADCATIA

Design datamanagement

Manufacturing follow-up

Installationfollow-up

Maintenancemanagement

Data publishing

Workflow actions

13

Credit: D. Widegren

Page 14: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

PLM @ CERN in numbersDocument & Drawings (incl. CAD):

~1,500.000 documents & drawings

~7,000 new documents & drawings created per month

Components:

~1,300,000 registered individually followed equipment

~3,000,000 equipment interventions/jobs logged

~ 15,000 equipment interventions/jobs logged per month

14

Credit: D. Widegren

Page 15: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

15

CASTOR and Oracle, tapes• Home made mass storage system, relies on Oracle

databases for name server, request handling and staging

• 4 libraries, SL8500• 10088x4 = 40K slots (4500 free)• Occupancy: 65PB worth of data• Drives: 20 T10KB legacy drives; 40 T10KC drives (to be

replaced by T10KD's)

Credit: German Cancio Melia

Page 16: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

CERN Disk/Tape Storage Management @ storage-day.ch

CASTOR Archive in Numbers

16

Data:• ~90PB of data on tape; 250M files• Up to 4.5 PB new data per month• Over 10GB/s (R+W) peaks

Credit: German Cancio Melia

Page 17: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

17

Experiment online systems• Experiments rely on

a SCADA system for their control

• Up to 150,000 changes / second stored in Oracle databases

Page 18: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

18

Experiment offline systems• Geometry DB

• Conditions DB

Credit: Vakho Tsulaia

Page 19: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

19

And… Fidelio Micros for CERN hotel

Page 20: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

20

Oracle at CERN• From accelerator control to

• accelerator logging, • administration, • engineering systems, • access control, • laboratory infrastructure (cabling,

network configuration, etc.), • mass storage system,• experiment online systems,• experiment offline systems,• CERN hotel• Etc.

Page 21: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

21

CERN openlab• Public-private partnership between CERN and

leading ICT companies, currently in fourth phase (started in 2003).

• Its mission is to accelerate the development of cutting-edge solutions to be used by the worldwide LHC community.

• Innovative ideas aligned between CERN and the partners.

• “Fellows” and summer students

Page 22: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.
Page 23: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

23

CERN• European Organization for Nuclear Research

• Founded in 1954• Research: Seeking and finding answers to questions about the Universe • Technology, International collaboration, Education

Twenty one Member StatesAustria, Belgium, Bulgaria, Czech Republic, Denmark, Finland, France, Germany, Greece, Italy, Israel, Hungary, Netherlands, Norway, Poland, Portugal, Slovakia, Spain, Sweden, Switzerland, United Kingdom

Seven Observer StatesEuropean Commission, USA, Russian Federation, India, Japan, Turkey, UNESCO

People~2400 Staff, ~900 Students, post-docs and undergraduates, ~9000 Users,~2000 Contractors

Associate Member StatesSerbia

Candidate StateRomania

Page 24: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

World’s largest computing grid - WLCG

•1 PB raw data per second before filtering / >20 PB of new data annually •68,889 physical CPUs / 305,935 logical CPUS•157 computer centres around the world

24

Page 25: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

25

Oracle at CERN, version 2.3

Credit: N. Segura Chinchilla

Page 26: 13 October 2014 Eric Grancher, head of database services, CERN IT Manuel Martin Marquez, data scientist, CERN openlab.

LHC

17 miles (27km) long tunnel

Thousands of superconducting magnets

Coldest place in the Universe: 1.9 degrees kelvin

Ultra vacuum: 10x emptier than on the Moon

600 million collisions per second

• The largest particle accelerator & detectors