Transcript Conception et optimisation d`une base de données

Conception et optimisation d’une base de
données relationnelle pour la
configuration d’expériences HEP
Implementation and optimization of a
relational database to configure and
manage HEP equipment
Lana Abadie
1
Sections
•
•
•
1)
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
Acknowledgments
Abstract
General introduction : context of the problem
State of the art
LHCb experiment
Introduction to databases
Requirements and use cases
Data table schema
Integration in the LHCb environment
Validation
Conclusions
Bibliography
French summary
annexes
Lana Abadie
2
1) State of the art
1.1 HEP experiments : what are the main characteristics
1.2 Configuring Equipment of an experiment : what are the
different steps? What are the specific concepts?
1.3 Need to use DB technology
1.4 Solutions in the past : ALEPH & DELPHI
1.5 Current possible solution : CMS & ATLAS
1.6 Use of conf DB in other fields
Lana Abadie
3
2) LHCb experiment
2.1 LHCb overview : Give the objectives, dataflow.
2.2 The detector : what are the different parts?
2.3 The online system: what is it?
2.4 LHCb Equipment management : what does it consist of?
2.5 Scope of investigation : what are the boundaries of my
work?
2.6 Thesis objectives : what are my objectives?
Lana Abadie
4
3) Introduction to databases
3.1 DBMS : what is it? What are the different types?
3.2 ERM : what does it consist of?
3.3 SQL: what is it?
3.4 Oracle : how does it work?
3.5 PL/SQL : what is it?
3.6 Database interfaces: what are the different types?
Lana Abadie
5
4) Requirements and use cases
4.1 Requirements. What are the requirements?
4.2 Use case 1 TFC. How will the TFC use CIC DB?
4.3 Use case 2 sub-detectors. How will the sub-detectors use
CIC DB? (ones which are different from the TFC)
4.4 Use case 3 the DAQ. How will the DAQ use CIC DB?
4.5 Use case 4 Inventory and history. What do the different
subsystems expect as information?
4.6 Conclusion
Lana Abadie
6
5) Data table schema
5.1 Introduction. Why did I choose the ERM?
5.2 Configuration representation. How did IT-JCOP represent
configuration data?
5.3 Inventory and history design : How did I model this info?
5.4 Connectivity design. How did I model this info?
5.5 Routing tables and paths. How did I model paths using
information from the CIC DB?
5.6 Conclusion
Lana Abadie
7
6) Integration in the LHCb environment
6.1 JCOP ConfDB tool : what are the features of this
tool?
6.2 A C-library to query connectivity & inventory data :
what are the different functions? How was it implemented?
6.3 CDBVis : What are the different features of this tool?
6.4 Security and concurrency and consistency. What about
security issues? How did I ensure consistency updates?
Are my queries safe enough when concurrent accesses?
6.6 Conclusion
Lana Abadie
8
7) Validation
6.1 Validity of the table schema (connectivity + inventory)
6.2 Benchmarks to test queries against a prototype of CIC
DB (connectivity and inventory info). Simulation of multiuser environment (stress testing).
Lana Abadie
9
8) Conclusions
8.1 Project status : were the aims and objectives satisfied?
8.2 Future work : ?
Lana Abadie
10