A  B  C  D  E  F  G  H  I  J  K  L  M  N  O  P  Q  R  S  T  U  V  W  X  Y  Z  special characters  sybase-tech-blog

  Bookmark us:
  
  
  RSS:
  Add to Google
  translated in:
   
Web en.sybinfo.org

what's new (latest 20)
UnpinUnpin
sybmonsybmon
dbcc dbrebootdbcc dbreboot
Pluggable Component Interface (PCI)Pluggable Component Interface
keep countkeep count
lct_admin("lastchance")lct_admin("lastchance&quo
sp_displayrolessp_displayroles
sp_activerolessp_activeroles
locking schemelocking scheme
Stored ProceduresStored Procedures
TRIGGER EVENTTRIGGER EVENT
Virtual Page SizeVirtual Page Size
ALTER EVENTALTER EVENT
Joins: StrategiesJoins: Strategies
Join OperatorsJoin Operators
SubqueriesSubqueries
sa_enable_auditing_typesa_enable_auditing_type
sa_disable_auditing_typesa_disable_auditing_type
AuditingAuditing
sa_audit_stringsa_audit_string
internal
goto: contact
goto: disclaimer


Category: Definition

Relational Database

A relational database is a database, based on the relational database model, which was first suggested by Edgar F. Codd in 1970. In this model, the term "relation" is described in the strictly mathematical sense of the "terminus technicus". In essence, it is a mathematical model for a table. Data is maintained in form of two-dimensional tables, which can be joined via keys, i.e. via primary keys or foreign keys. . The majority of database systems are designed for relational databases. This is the reason why in general the term "database" is used for relational databases or relational database systems.

Relational databases can be queried using SQL. In ASE T-SQL is used.

Usually the Entity-Relationship-Model, or versions of it, is used to design relational databases. It serves to create a conceptual schema, which can be implemented using a DBMS . This step is called logical concept or data model and results in a database schema in the implementation data model of the DBMS.

In earlier days hierarchical database systems and network-based database systems were used for corporate data processing. In a few special cases these are still in use today.

Basic Rules for a Relational Database according to Edgar F. Codd

  • Each relation is a two-dimensional table and corresponds to a type of relation
  • Every row of this relation (table) is called tuple and describes a precise tuple of the type of relation depicted by the relation (table).
  • Each column of the relation (table) correponds to a attribute of the relational type. The precise tuple are thus described by their respective attribute values.
  • The number of attributes depicts the degree of a relation
  • The cardinality of a relation is equivalent to ist number of tuples
  • If an attribute can have only a specific number of values, the aggregation of all attribute values is called domain (range of values)
  • It is not relevant in which order rows and columns are stored in a table
  • Having two identical columns is not valid
  • values of attributes are atomic

Important Terms of Relational Databases

Table
A table is an aggregation of a number of records with the same structure, comparable to a filing box. In the DBMS these records are usually displayed in a tabular form. A table for addresses could for example store last name, first name, street, house number, place and phone number of a user.

Data Set, Record, Tuple
The information making up a record, data set or tuple depicts one row in a table. The record can be displayed as a row or on a web page in a clearly arranged way. A record is comparable to one card in a filing box.

Data Field, Field, Attribute
A data field or attribute is part of a record. For example, the field with names in a table with addresses. Fields can have data of different kind but they must all be of the same datatype. A field is comparable to one row on one card of a filing box. A distinction is made here between key attributes and non-key attributes, i.e. all other attributes. The key attribute is used to identify and join records. All other attributes contain data which is independent of the key. For example, "personnel number" is the key to which the attributes "entry date" or "date of birth" can be assigned.

Query
A query is used to display the contents or part of the contents of one or more tables. The result is a (temporary) table filtered according to specific criteria. Compared to a filing box, a query corrsponds to a selection of certain cards according to specific criteria. For example, one could match all cards with customers whose name starts with A and all cards of the same customers with unpaid bills. Simple queries are commands for alphabetic sorting or filtering according to specific criteriaBeim Vergleich mit einem Karteikasten entspricht eine Abfrage der Auswahl einiger Karten nach bestimmten Kriterien. Queries are don in SQL or T-SQL.

View
A View Can be described as a stored query of one or more joined tables. For users a view behaves just like a table. Views are created for structured queries. This way it is easy to restrict user access to specific records. Because DMLs cannot be used with views, malicious or unintentional manipulation of data records can be avoided.

Report
Reports are formatted summaries of one or more queries, sometimes even preprocessed for printing. Reports can be generated by tools or software, delivered by the vendor, third-party-vendors or integrated into the DBMS. These report generators are not part of the core DBMS, but will normally be provided as an add-on by the vendor or third-party-vendors.