Chapter 6 Physical Database Design and Performance Modern

Chapter 6: Physical Database Design and Performance Modern Database Management 6 th Edition Jeffrey A. Hoffer, Mary B. Prescott, Fred R. Mc. Fadden © Prentice Hall, 2002 1

The Physical Design Stage of SDLC (figures 2. 4, 2. 5 revisited) Purpose –develop technology specs Deliverable – pgm/data structures, technology purchases, organization redesigns Project Identification and Selection Project Initiation and Planning Analysis Logical Design Physical Design Database activity – physical database design Implementation Maintenance Chapter 6 © Prentice Hall, 2002 2

Physical Database Design l Purpose - translate the logical description of data into the technical specifications for storing and retrieving data l Goal - create a design for storing data that will provide adequate performance and insure database integrity, security and recoverability Chapter 6 © Prentice Hall, 2002 3

Physical Design Process Inputs l. Normalized l. Volume relations definitions l. Response time expectations security needs l. Backup/recovery l. Integrity l. DBMS l. Attribute estimates l. Attribute l. Data Decisions needs expectations technology used Chapter 6 data types l. Physical record descriptions (doesn’t always match logical design) Leads to l. File organizations l. Indexes and database architectures l. Query © Prentice Hall, 2002 optimization 4

Figure 6. 1 - Composite usage map (Pine Valley Furniture Company) Chapter 6 © Prentice Hall, 2002 5

Figure 6. 1 - Composite usage map (Pine Valley Furniture Company) Data volumes Chapter 6 © Prentice Hall, 2002 6

Figure 6. 1 - Composite usage map (Pine Valley Furniture Company) Access Frequencies (per hour) Chapter 6 © Prentice Hall, 2002 7

Figure 6. 1 - Composite usage map (Pine Valley Furniture Company) Usage analysis: 200 purchased parts accessed per hour 80 quotations accessed from these 200 purchased part accesses 70 suppliers accessed from these 80 quotation accesses Chapter 6 © Prentice Hall, 2002 8

Figure 6. 1 - Composite usage map (Pine Valley Furniture Company) Usage analysis: 75 suppliers accessed per hour 40 quotations accessed from these 75 supplier accesses 40 purchased parts accessed from these 40 quotation accesses Chapter 6 © Prentice Hall, 2002 9

Designing Fields l Field: smallest unit of data in database l Field design – Choosing data type – Coding, compression, encryption – Controlling data integrity Chapter 6 © Prentice Hall, 2002 10

Choosing Data Types l CHAR – fixed-length character l VARCHAR 2 – variable-length character (memo) l LONG – large number l NUMBER – positive/negative number l DATE – actual date l BLOB – binary large object (good for graphics, sound clips, etc. ) Chapter 6 © Prentice Hall, 2002 11

Figure 6. 2 Example code-look-up table (Pine Valley Furniture Company) Code saves space, but costs an additional lookup to obtain actual value. Chapter 6 © Prentice Hall, 2002 12

Field Data Integrity l l Default value - assumed value if no explicit value Range control – allowable value limitations (constraints or validation rules) Null value control – allowing or prohibiting empty fields Referential integrity – range control (and null value allowances) foreign-key to primarykey match-ups (in a same or different table) Chapter 6 © Prentice Hall, 2002 13

Handling Missing Data l Substitute an estimate of the missing value (e. g. using a formula) l Construct a report listing missing values. – Can be done by setting up a trigger in database definition l In programs, ignore missing data unless the value is significant l. Triggers Chapter 6 can be used to perform these operations © Prentice Hall, 2002 14

Example Missing Data l Customer zip code is null. A report summarizes total sales by month and zip code. l How should sales to customers with unknown zip codes be handles? l Two option: – Use default value – Use not permitting null value Chapter 6 © Prentice Hall, 2002 15

Physical Records l Physical Record: A group of fields stored in adjacent memory locations and retrieved together as a unit by DBMS l Page: The amount of data read or written in one I/O operation (OS) l Blocking Factor: The number of physical records per page Chapter 6 © Prentice Hall, 2002 16

Denormalization Transforming normalized relations into unnormalized physical record specifications l Benefits: l – Can improve performance (speed) be reducing number of table lookups (i. e reduce number of necessary join queries) l Costs (due to data duplication) – Wasted storage space – Data integrity/consistency threats l Common denormalization opportunities (combining tables to avoid doing joins) – One-to-one relationship (Fig 6. 3) – Many-to-many relationship with attributes (Fig. 6. 4) – Reference data (1: N relationship where 1 -side has data not used in any other relationship) (Fig. 6. 5) Chapter 6 © Prentice Hall, 2002 17

Fig 6. 5 – A possible denormalization situation: reference data Reference data exist in one side and no participation to other database. Extra table access required Data duplication Chapter 6 © Prentice Hall, 2002 18

Partitioning l Horizontal Partitioning: Distributing the rows of a table into several separate files – Useful for situations where different users need access to different rows – Make sense when different categories of row are processed separately – Three types: Key Range Partitioning, Hash Partitioning, or Composite Partitioning l Vertical Partitioning: Distributing the columns of a table into several separate files – Useful for situations where different users need access to different columns – The primary key must be repeated in each file l Combinations of Horizontal and Vertical Partitions often correspond with User Schemas (user views) Chapter 6 © Prentice Hall, 2002 19

Example Customer relation could be broken into four regional customer based on the value of a field region l All customer can be reconstructed by SQL UNION operator. l Vertical partitioning l – Break columns of Part relation into – Engineering related part data – Accounting related part data – Sales rlated part data Chapter 6 © Prentice Hall, 2002 20

Partitioning l Advantages of Partitioning: – Records used together are grouped together (efficiency) – Each partition can be optimized for performance (local – – l optimization) Security, different access grant to different user Recovery Partitions stored on different disks: contention Take advantage of parallel processing capability Disadvantages of Partitioning: – Slow retrievals across partitions – Complexity Chapter 6 © Prentice Hall, 2002 21

Data Replication l l l Purposely storing the same data in multiple locations of the database Improves performance by allowing multiple users to access the same data at the same time with minimum contention Sacrifices data integrity due to data duplication Best for data that is not updated often Worth while a table accessed frequently and the cost of extra storage & maintenance are not great. Chapter 6 © Prentice Hall, 2002 22

Designing Physical Files l Physical File: – A named portion of secondary memory allocated for the purpose of storing physical records l Constructs to link two pieces of data: – Sequential storage. – Pointers. l File Organization: – How the files are arranged on the disk. – In modern RDBMS user do not have to design file organization but may be allowed to select a fileorganization and its parameter. l Access Method: – How the data can be retrieved based on the file Chapter 6 organization. © Prentice Hall, 2002 23

DATA FILE DAN TABLESPACE Configuration file Database Data file Log file Arsitektur fisik database Arsitektur database secara logic Memisahkan database atas Table Spaces. • Setiap database memiliki 1 table Space. • Setiap tablespace memiliki nol atau lebih segment. • Setiap Segmen memiliki 1 atau lebih extents. • Setiap Extent memiliki 1 atau lebih Database Block Tablespace Segment Extent Db-Block OS-Block adalah blok data yang didefinisikan OS, yaitu yang merelasikan antara DB-Blok dengan Data File. Arsiktektur Logic databse 24

Figure 6 -7 (a) Sequential file organization l Records of the file are stored in sequence by the primary key field values. 1 2 If sorted – every insert or delete requires resort If not sorted Average time to find desired record = n/2. n Chapter 6 © Prentice Hall, 2002 25

Indexed File Organizations Index – a separate table that contains organization of records for quick retrieval l Primary keys are automatically indexed l Oracle has a CREATE INDEX operation, and MS ACCESS allows indexes to be created for most field types l Indexing approaches: l – – B-tree index, Fig. 6 -7 b Bitmap index, Fig. 6 -8 Hash Index, Fig. 6 -7 c Join Index, Fig 6 -9 Chapter 6 © Prentice Hall, 2002 26

Fig. 6 -7 b – B-tree index Indexes can be built on top of indexes. Leaves of the tree are all at same level consistent access time uses a tree search Average time to find desired record = depth of the tree Chapter 6 © Prentice Hall, 2002 27

Fig 6 -7 c Hashed file or index organization Hash algorithm Usually uses divisionremainder to determine record position. Records with same position are grouped in lists. Chapter 6 © Prentice Hall, 2002 28

Fig 6 -8 Bitmap index organization Chapter 6 Bitmap saves on space requirements Rows - possible values of the attribute Columns - table rows Bit indicates whether the attribute of a row has the values © Prentice Hall, 2002 29

Fig 6 -9 Join Index – speeds up join operations Chapter 6 © Prentice Hall, 2002 30

Join Index l Without join index, in fig 6. 9 a, any query to find stores and customers in the same city would have to compute the equivalent of the join index each time the query is run. l For very large tables, joining all rows in one table with matching rows in another can be very time consuming and significant delay in-online query. Chapter 6 © Prentice Hall, 2002 31

Clustering Files In some relational DBMSs, related records from different tables can be stored together in the same disk area l Useful for improving performance of join operations l Primary key records of the main table are stored adjacent to associated foreign key records of the dependent table l e. g. Oracle has a CREATE CLUSTER command (see p. 231) l Chapter 6 © Prentice Hall, 2002 32

Rules for Using Indexes 1. Use on larger tables 2. Index the primary key of each table 3. Index search fields (fields frequently in WHERE clause) 4. Fields in SQL ORDER BY and GROUP BY commands 5. When there are >100 different values but not when there are <30 values Chapter 6 © Prentice Hall, 2002 33

Rules for Using Indexes 6. DBMS may have limit on number of indexes per table and number of bytes per indexed field(s) 7. Null values will not be referenced from an index 8. Use indexes heavily for non-volatile databases; limit the use of indexes for volatile databases Why? Because modifications (e. g. inserts, deletes) require updates to occur in index files Chapter 6 © Prentice Hall, 2002 34

RAID l Redundant Array of Inexpensive Disks l A set of disk drives that appear to the user to be a single disk drive l Allows parallel access to data (improves access speed) l Pages are arranged in stripes Chapter 6 © Prentice Hall, 2002 35

Figure 6 -10 – RAID with four disks and striping Here, pages 1 -4 can be read/written simultaneously Chapter 6 © Prentice Hall, 2002 36

Raid Types (Figure 6 -11) l Raid 0 – – l l Maximized parallelism No redundancy No error correction no fault-tolerance Error correction in one disk – Record spans multiple data disks (more than RAID 2) – Not good for multi-user environments, – l Raid 1 Error correction in one disk – Multiple records per stripe – Parallelism, but slow updates due to error correction contention Raid 2 – No redundancy – One record spans across data disks – Error correction in multiple disks – reconstruct damaged data Chapter 6 Raid 4 – – Redundant data – fault tolerant – Most common form l Raid 3 l Raid 5 Rotating parity array ‒ Error correction takes place in same disks as data storage ‒ Parallelism, better performance than Raid 4 ‒ © Prentice Hall, 2002 37

(figure 6 -12 Database Architectures Legacy Systems Chapter 6 Current Technology Data Warehouses © Prentice Hall, 2002 38

Query Optimization l Parallel Query Processing l Override Automatic Query Optimization l Data Block Size -- Performance tradeoffs: – Block contention – Random vs. sequential row access speed – Row size – Overhead l Balancing Chapter 6 I/O Across Disk Controllers © Prentice Hall, 2002 39

Query Optimization l Wise use of indexes l Compatible data types l Simple queries l Avoid query nesting l Temporary tables for query groups l Select only needed columns l No sort without index Chapter 6 © Prentice Hall, 2002 40