Lecture 1: Overview of CSCI 585LogisticsPre-ReqInstructor DetailsTAOutlineDatabase Management Systems (DBMS)DATABASE & DBMSBEFORE DBMSAFTER DBMSWHY A DBMS?DBMS ARCHITECTUREAn Emerging PhenomenaExampleWHAT HAS CHANGED?Cross-roads585 Syllabus585 Syllabus (Cont…)Slide 19Slide 20Slide 21Slide 22Slide 23GradingFor next lectureLecture 1: Overview of CSCI 585Lecture 1: Overview of CSCI 585Prof. Shahram GhandeharizadehProf. Shahram GhandeharizadehDirector of USC Database Lab (http://dblab.usc.edu)Director of USC Database Lab (http://dblab.usc.edu)Computer Science DepartmentComputer Science DepartmentUniversity of Southern CaliforniaUniversity of Southern CaliforniaLogisticsLogisticsCollection of technical papers:Collection of technical papers:ACM/IEEE/Springer digital libraries.ACM/IEEE/Springer digital libraries.URLs work from USC machines.URLs work from USC machines.Pre-req for the course:Pre-req for the course:CSCI 485: Introduction to File and Database CSCI 485: Introduction to File and Database Management, andManagement, andKnowledge Knowledge C++ programming language.C++ programming language.Extensive use of Blackboard for homework Extensive use of Blackboard for homework and project submissions. Make sure to have and project submissions. Make sure to have access to:access to:http://den.usc.eduhttp://den.usc.eduPower-point of presentations also available Power-point of presentations also available from http://dblab.usc.edufrom http://dblab.usc.eduPre-Req585 assumes you know the following:Transactions and their ACID properties.Concurrency control protocols such as locking and time-stamp based protocols.Crash recovery techniques such as logging and shadow paging.Physical characteristics of magnetic disks.SQLRelational algebra operatorsER data modeling Alternative normal forms.Visit http://dblab.usc.edu/csci485 for an overview of this material.Instructor DetailsInstructor DetailsDr. Shahram GhandeharizadehDr. Shahram GhandeharizadehOffice: SAL 208Office: SAL 208E-mail: [email protected]: [email protected]: 213-740-4781Phone: 213-740-4781Office Hours:Office Hours:Tuesday: 12:30 to 2 pmTuesday: 12:30 to 2 pmThursday: 4:30 to 5:30 pmThursday: 4:30 to 5:30 pmClass URL: http://dblab.usc.edu/csci585Class URL: http://dblab.usc.edu/csci585TATAShahin ShayandehShahin ShayandehOffice: SAL 200COffice: SAL 200CE-mail: [email protected]: [email protected] Hours:Office Hours:Mondays: 3:30 to 5 pmMondays: 3:30 to 5 pmThursday: 12:30 to 2 pmThursday: 12:30 to 2 pmOutlineOutlineMotivation for DBMSMotivation for DBMSAn outline for the course materialAn outline for the course materialGrading: Assignments and projectsGrading: Assignments and projectsDatabase Management Database Management Systems (DBMS)Systems (DBMS)Used almost on a daily basis for either Used almost on a daily basis for either individual or business use.individual or business use.Relational database vendors were one Relational database vendors were one of the fastest growing sectors during of the fastest growing sectors during the .COM boom!the .COM boom!DATABASE & DBMSDATABASE & DBMSDatabaseDatabase: An integrated collection of : An integrated collection of data, usually stored on secondary data, usually stored on secondary storage, typically describing the storage, typically describing the activities of one or more related activities of one or more related organizations.organizations.Database management systemDatabase management system ((DBMSDBMS)): : A collection of software/programs A collection of software/programs designed to assist in maintaining and designed to assist in maintaining and utilizing large collections of data.utilizing large collections of data.BEFORE DBMSBEFORE DBMSDataDataUser 1User 2Application programsApplication programsData managed by DBMSAFTER DBMSAFTER DBMSUser 1User 2DBMSApplication programsApplication programsWHY A DBMS?WHY A DBMS?1.1.Reduced application development timeReduced application development time2.2.Data independence: Application programs not dependent on Data independence: Application programs not dependent on data representation and storage detailsdata representation and storage details3.3.Data sharing: data is better utilized (discovered and reused), Data sharing: data is better utilized (discovered and reused), redundancy of data is minimizedredundancy of data is minimized4.4.Data integrity and consistency: one may enforce consistency Data integrity and consistency: one may enforce consistency constraints on data, e.g., number of seats sold ≤ number of seats constraints on data, e.g., number of seats sold ≤ number of seats on the plane × 1.1on the plane × 1.15.5.Centralized control: DBA tunes the database to balance user's Centralized control: DBA tunes the database to balance user's needsneeds6.6.Security: mechanisms to prevent unauthorized access. These Security: mechanisms to prevent unauthorized access. These mechanisms are based on content instead of file-oriented mechanisms are based on content instead of file-oriented approach.approach.7.7.Concurrency control: avoids undesirable race conditions that Concurrency control: avoids undesirable race conditions that arise with simultaneous access/updates to dataarise with simultaneous access/updates to data8.8.Crash recovery: ensures the integrity of data in the presence of Crash recovery: ensures the integrity of data in the presence of failuresfailuresDBMS ARCHITECTUREDBMS ARCHITECTUREUser 1User n…Conceptual schemaConceptual schemaPhysical dataDBDBMSData managed by DBMSAn Emerging PhenomenaAn Emerging PhenomenaUser 1User 2DBMSApplication programsApplication programsExampleF. Chang et. al. Bigtable: A Distributed Storage System for Structured Data. In OSDI 2006. Last paragraph of the paper:“Finally, we have found that there are significant advantages to building our own storage solution at Google. We have gotten substantial amount of flexibility from designing our own data model for Bigtable. In addition, our control over Bigtable’s implementation, and the other Google infrastructure upon which Bigtable depends, means that we can remove bottlenecks and inefficiencies as they arise.”WHAT HAS CHANGED?WHAT HAS CHANGED?1.1.Relational database technology is now more than a quarter of Relational database technology is now more than a quarter of century old.century old.2.2.While concepts such