Storage and File Organization File Organization Basics A
Storage and File Organization
File Organization § Basics § A database is a collection of files, § A file is a collection of records § A record (tuple) is a collection of fields (attributes) § Files are stored on Disks (that use blocks to read and write) § Two important issues: 1. Representation of each record 2. Grouping/Ordering of records and storage in blocks Database System Concepts 11. 2
File Organization § Goal and considerations: § Compactness § Overhead of insertion/deletion § Retrieval speed § sometimes we prefer to bring more tuples than necessary into MM and use CPU to filter out the unnecessary ones! Database System Concepts 11. 3
Record Representation § Fixed-Length Records § Example Account( acc-number char(10), branch-name char(20), balance real) Each record is 38 bytes. Store them sequentially, one after the other Record 1 at position 0, record 2 at position 38, record 3 at position 76 etc Compactness (350 bytes) Database System Concepts 11. 4
Fixed-Length Records § Simple approach: § Store record i starting from byte n (i – 1), where n is the size of each record. § Record access is simple but records may cross blocks § Modification: do not allow records to cross block boundaries § Insertion of record i: Add at the end § Deletion of record i: Two alternatives: § move records: 1. i + 1, . . . , n to i, . . . , n – 1 2. Move record n to position i § do not move records, but link all free records on a free list Database System Concepts 11. 5
Free Lists § 2 nd approach: FLR with Free Lists § § § Store the address of the first deleted record in the file header. Use this first record to store the address of the second deleted record, and so on Can think of these stored addresses as pointers since they “point” to the location of a record. Better handling ins/del Less compact Database System Concepts 11. 6
Variable-Length Records § 3 rd approach: Variable-length records arise in database systems in several ways: § Storage of multiple record types in a file. § Record types that allow variable lengths for one or more fields. § Record types that allow repeating fields or multivalued attribute. § Byte string representation § Attach an end-of-record ( ) control character to the end of each record § Difficulty with deletion (leaves holes) § Difficulty with growth 4 Field Count Database System Concepts R 1 R 2 11. 7 R 3
Variable-Length Records: Slotted Page Structure § 4 th approach VLR-SP § Slotted page header contains: § number of record entries § end of free space in the block § location and size of each record § Records stored at the bottom of the page § External tuple pointers point to record ptrs: § rec-id = <page-id, slot#> Database System Concepts 11. 8
Rid = (i, N) Page i Rid = (i, 2) Rid = (i, 1) 20 N 16. . . 24 2 N 1# slots SLOT DIRECTORY Insertion: 1) Use Free Space Pointer (FP) to find space and insert 2) Find available ptr in the directory (or create a new one) 3) adjust FP and number of records Deletion ? Database System Concepts 11. 9 Pointer to start of free space
Variable-Length Records (Cont. ) § Fixed-length representation: § reserved space § pointers § 5 th approach: Fixed Limit Records (for VLR) § Reserved space – can use fixed-length records of a known maximum length; unused space in shorter records filled with a null or end-of-record symbol. Database System Concepts 11. 10
Pointer Method § § 6 th approach: Pointer method § A variable-length record is represented by a list of fixed-length records, chained together via pointers. § Can be used even if the maximum record length is not known Database System Concepts 11. 11
Pointer Method (Cont. ) § Disadvantage to pointer structure; space is wasted in all records except the first in a chain. § Solution is to allow two kinds of block in file: § Anchor block – contains the first records of chain § Overflow block – contains records other than those that are the first records of chains. Database System Concepts 11. 12
Ordering and Grouping records § Issue #1: § In what order we place records in a block? 1. Heap technique: assign anywhere there is space 2. Ordered technique: maintain an order on some attribute So, we can use binary search if selection on this attribute. Database System Concepts 11. 13
Sequential File Organization § Suitable for applications that require sequential processing of the entire file § The records in the file are ordered by a search-key Database System Concepts 11. 14
Sequential File Organization (Cont. ) § Deletion – use pointer chains § Insertion –locate the position where the record is to be inserted § if there is free space insert there § if no free space, insert the record in an overflow block § In either case, pointer chain must be updated § Need to reorganize the file from time to restore sequential order Database System Concepts 11. 15
Clustering File Organization § Simple file structure stores each relation in a separate file § Can instead store several relations in one file using a clustering file organization § e. g. , clustering organization of customer and depositor: SELECT account-number, customer-name FROM depositor d, account a WHERE d. customer-name = a. customer-name Ø good for queries involving depositor account, and for queries involving one single customer and his accounts Ø bad for queries involving only customer Ø results in variable size records Database System Concepts 11. 16
File organization § Issue #2: In which blocks should records be placed Many alternatives exist, each ideal for some situation , and not so good in others: § Heap files: Add at the end of the file. Suitable when typical access is a file scan retrieving all records. § Sorted Files: Keep the pages ordered. Best if records must be retrieved in some order, or only a `range’ of records is needed. § Hashed Files: Good for equality selections. Assign records to blocks according to their value for some attribute Database System Concepts 11. 17
Data Dictionary Storage Data dictionary (also called system catalog) stores metadata: that is, data about data, such as § Information about relations § § § names of relations names and types of attributes of each relation names and definitions of views integrity constraints User and accounting information, including passwords Statistical and descriptive data § number of tuples in each relation § Physical file organization information § How relation is stored (sequential/hash/…) § Physical location of relation § operating system file name or § disk addresses of blocks containing records of the relation § Information about indices Database System Concepts 11. 18
Data dictionary storage § Stored as tables!! § E-R diagram? § Relations, attributes, domains § Each relation has name, some attributes § Each attribute has name, length and domain § Also, views, integrity constraints, indices § User info (authorizations etc) § statistics Database System Concepts 11. 19
A-name 1 relation position N has attribute domain Database System Concepts 11. 20
Data Dictionary Storage (Cont. ) § A possible catalog representation: Relation-metadata = (relation-name, number-of-attributes, storage-organization, location) Attribute-metadata = (attribute-name, relation-name, domain-type, position, length) User-metadata = (user-name, encrypted-password, group) Index-metadata = (index-name, relation-name, index-type, index-attributes) View-metadata = (view-name, definition) Database System Concepts 11. 21
- Slides: 21