PD.Relational Data Management: Difference between revisions

From OIAr
Jump to navigation Jump to search
No edit summary
No edit summary
 
(20 intermediate revisions by the same user not shown)
Line 1: Line 1:
{{Maturity|0}}
{{Maturity|0
|maturity=0
}}
{{Pageheaderbox4PD
{{Pageheaderbox4PD
|name=Relational Data Management
|name=Relational Data Management
Line 12: Line 14:
* a "transaction mechanism", to enable interaction with other (instances of) data management facilities and applications.
* a "transaction mechanism", to enable interaction with other (instances of) data management facilities and applications.
These elements are provided by the [[DOT.Data|Data]] [[FD.Engine|Engine]]. Data objects are stored using a predefined (structured) [[DOT.Data|Data]] [[FD.Store|Store]]. Both functions form the core of this Pattern.
These elements are provided by the [[DOT.Data|Data]] [[FD.Engine|Engine]]. Data objects are stored using a predefined (structured) [[DOT.Data|Data]] [[FD.Store|Store]]. Both functions form the core of this Pattern.
{{Pattern Realizes
{{Pattern Realizes
|service=SD.Application Database
|service=SD.Application Database
}}
}}
{{Pattern Graphic
{{Pattern Definition Graphic
|graphic=PD.Relational Data Management.png
|graphic=PD.Relational Data Management.png
|source=PD.
|size=800px
|size=800px
|title=Relational Data Management pattern
|title=Relational Data Management pattern
}}
}}
{{Generic Pattern Composition}}
{{Pattern Definition Composition
|Row_dataobjectype=DOT.Traffic
}}
{{Pattern Definition Composition Row
{{Pattern Definition Composition Row
|function=FD.Engine
|function=FD.Engine
|dataobjectype=DOT.Data
|dataobjecttype=DOT.Data
|choice=Must
|choice=Must
|reason=The Data Engine is the heart of the pattern, as it offers the intelligence to control strictly structured data handling.
|reason=The Data Engine is the heart of the pattern, as it offers the intelligence to control strictly structured data handling.
Line 31: Line 33:
{{Pattern Definition Composition Row
{{Pattern Definition Composition Row
|function=FD.Store
|function=FD.Store
|dataobjectype=DOT.Table
|dataobjecttype=DOT.Table
|choice=Must
|reason=The Table Store offers the predefined storage structure. This is one of the key characteristics of relational data management.
}}
{{Pattern Definition Composition Row
|function=FD.Logging
|dataobjecttype=DOT.Transaction
|choice=Must
|choice=Must
|reason=The Table Store offers the predefined storage structure that is one of the key characteristics of relational data management.
|reason=Keeping a record of transactions, in order to be able to analyze database interaction and to revert and replay transactions.
}}
{{Pattern Definition Composition Row
|function=FD.Archive
|dataobjecttype=DOT.Log
|choice=May
|reason=Preservation of transaction logs, in order to be able to get further back in time regarding transactions.
}}
}}
{{Pattern Definition Composition Row
{{Pattern Definition Composition Row
|function=FD.Encryption
|function=FD.Encryption
|dataobjectype=DOT.Traffic
|dataobjecttype=DOT.Traffic
|choice=May
|choice=May
|reason=When data is exchanged over unsecured data transport facilities, it might be a good measure to encrypt the data at engine level.
|reason=When data is exchanged over unsecured data transport facilities, it might be a good measure to encrypt the data at engine level.
}}
}}
{{Generic Pattern Composition Row
{{Pattern Definition Composition Row
|function=GF.Controlling
|function=FD.Caching
|choice=Must
|dataobjecttype=DOT.Transaction
|reason=When allowing consumers (such as users and/or administrators) the ability to change the configuration of the Data Management facility, special care may be required to prevent unauthorized access to data stored by the Pattern.
|choice=May
|reason=Retaining and keeping available of frequent issued transactions/queries, in order to improve performance.
}}
{{Pattern Definition Composition Row
|function=FD.Replication
|dataobjecttype=DOT.Transaction
|choice=May
|reason=Multiplying transactions over multiple database instances, to maintain exact copies on data engine level. Mostly this is done to create a higher availability
}}
{{Pattern Definition Composition Row
|function=FD.Replication
|dataobjecttype=DOT.Table
|choice=May
|reason=Multiplying tables over multiple database instances, to create an exact copy of a database on data engine level. This can be done to create a new development/test environment or to start creating a higher availability
}}
}}
{{Generic Pattern Composition Row
{{Pattern Definition Composition Row
|function=GF.Caching
|function=FD.Export
|dataobjecttype=DOT.File
|choice=May
|choice=May
|reason=Since the Data Management pattern underlies many other infrastructure facilities and applications, there often are requirements to guarantee a certain performance level, for which this Caching functionality may provide a significant boost. However, caching introduces additional challenges related to data being consistent and up-to-date, security risks for sensitive data residing in cache memory, et cetera.
|reason=Extraction of data from the database and putting this in a file. This is used to execute a bulk data transfer from one instance to another. It can be used to create a replica or to import a dataset from one database into the other (ETL). Because it does not require a connection between Data Engines, it offers more flexibility and security, compared to direct Data Engine interaction. In both cases, the proper interfaces should be available.
}}
}}
{{Table Ending}}
{{Pattern Definition Composition Row
{{Pattern Adjacent Services}}
|function=FD.Import
{{Generic Pattern Adjacent Service Row
|dataobjecttype=DOT.File
|service=GS.File Storage
|choice=May
|choice=Must
|reason=Loading of data from a file, to store this data in a database. This is used to restore a database from a replica or to import a large dataset from another database (ETL). Because it does not require a connection between Data Engines, it offers more flexibility and security, compared to direct Data Engine interaction. In both cases, the proper interfaces should be available.
|reason=Every Data Management pattern must store its data at a lower [[OIAm data view|data level]], which means it requires the use of a File Storage service. Even in commercial products that request access to Raw Storage this is true, as these products actually provide the File Storage functionality themselves.
}}
}}
{{Generic Pattern Adjacent Service Row
{{Pattern Definition Composition Row
|service=GS.Authentication+Authorization
|function=FD.Concealment
|dataobjecttype=DOT.Data
|choice=May
|choice=May
|reason=A suitable source for identity and permission information may be required for
|reason=Hiding of (privacy-sensitive) data objects when presenting/transferring/exporting data, by means of masking or scrambling.
* controlling access to the data stored in the Data Management facility;
* management of the data;
* managing the behaviour of the Data Management facility itself, using the Controlling functionality.
}}
}}
{{Generic Pattern Adjacent Service Row
{{Pattern Definition Composition Row
|service=GS.Data Protection Management
|function=FD.Tiering
|dataobjecttype=DOT.Table
|choice=May
|choice=May
|reason=The data being handled by this Pattern may need to be protected against data corruption and/or loss. Even though this might be delegated to the lower level of the [[GS.File_Storage|File Storage]] service that the Pattern uses, it's usually much better to handle data protection at the logical level of this Pattern, as at this level the interdependencies of parts of the datasets is known and controllable.
|reason=Making use of different storage tiers for table retention, to reduce costs.
}}
}}
{{Generic Pattern Adjacent Service Row
{{Pattern Definition Composition Row
|service=GS.Archive Management
|function=FD.Encryption
|dataobjecttype=DOT.Table
|choice=May
|choice=May
|reason=The Pattern may need to archive (parts of) a [[OIAm data view|structured data set]], and/or may require access to an archive containing (parts of) a structured data set, in order to fulfil requests from its clients; furthermore, it may be necessary to migrate (parts of) an active structured data set to an archive. For any of this, the Data Management pattern needs access to an Archive Management service.
|reason=Encrypting (making unreadable without a key) of data when stored in a structured data store.
}}
{{Pattern Definition Composition Row
|function=FD.Controlling
|dataobjecttype=DOT.Platform
|choice=Must
|reason=Administrative interface to configure and manage database systems.
}}
{{Pattern Definition Composition Row
|function=FD.Logging
|dataobjecttype=DOT.Platform
|choice=Must
|reason=Keeping track of events, activities, interactions that occur during database system operation.
}}
}}
{{Table Ending}}
{{Table Ending}}
{{Text Footer GP}}

Latest revision as of 20:35, 27 November 2019

Page maturity
This page has maturity level 0 (proposal)


This is a Pattern Definition document PD Relational Data Management Version: 0.3 OIAr logo
Document type: Pattern Definition Owner:

S.A.D. Jumelet


This Pattern provides the function layout to model "strictly structured data" handling, in a relational fashion.


Description

This Pattern Definition serves the creation of models in the Service Category "Data Management". This Pattern provides the function layout to model "strictly structured data" handling, i.e. it controls the creation, updating and querying of sets of organized data. It is called 'relational data management', because a schema is used to store data in a predefined way, with predefined relations between data objects. The provisioning of the following elements is important with this kind of data handling:

  • a "modeling language", to define schemas;
  • a "query language", to enable data object reading, creating, updating, and deleting;
  • a "transaction mechanism", to enable interaction with other (instances of) data management facilities and applications.

These elements are provided by the Data Engine. Data objects are stored using a predefined (structured) Data Store. Both functions form the core of this Pattern.

Services realized

This Pattern realizes the following service(s):

  • Application Database (This service provides a data store for applications with a predefined structure)

Functional and Integration view

This is the graphic representation of the function model of this Pattern Definition:

Relational Data Management pattern
Relational Data Management pattern


Pattern Definition Composition

This pattern is an aggregation of the following (mandatory and optional) functions, expressed in Data/Object Types and Fundtion Definitions:

Function Inclusion Description/Rationale
Data Engine recommended The Data Engine is the heart of the pattern, as it offers the intelligence to control strictly structured data handling.
Table Store recommended The Table Store offers the predefined storage structure. This is one of the key characteristics of relational data management.
Transaction Logging recommended Keeping a record of transactions, in order to be able to analyze database interaction and to revert and replay transactions.
Log Archive optional Preservation of transaction logs, in order to be able to get further back in time regarding transactions.
Traffic Encryption optional When data is exchanged over unsecured data transport facilities, it might be a good measure to encrypt the data at engine level.
Transaction Caching optional Retaining and keeping available of frequent issued transactions/queries, in order to improve performance.
Transaction Replication optional Multiplying transactions over multiple database instances, to maintain exact copies on data engine level. Mostly this is done to create a higher availability
Table Replication optional Multiplying tables over multiple database instances, to create an exact copy of a database on data engine level. This can be done to create a new development/test environment or to start creating a higher availability
File Export optional Extraction of data from the database and putting this in a file. This is used to execute a bulk data transfer from one instance to another. It can be used to create a replica or to import a dataset from one database into the other (ETL). Because it does not require a connection between Data Engines, it offers more flexibility and security, compared to direct Data Engine interaction. In both cases, the proper interfaces should be available.
File Import optional Loading of data from a file, to store this data in a database. This is used to restore a database from a replica or to import a large dataset from another database (ETL). Because it does not require a connection between Data Engines, it offers more flexibility and security, compared to direct Data Engine interaction. In both cases, the proper interfaces should be available.
Data Concealment optional Hiding of (privacy-sensitive) data objects when presenting/transferring/exporting data, by means of masking or scrambling.
Table Tiering optional Making use of different storage tiers for table retention, to reduce costs.
Table Encryption optional Encrypting (making unreadable without a key) of data when stored in a structured data store.
Platform Controlling recommended Administrative interface to configure and manage database systems.
Platform Logging recommended Keeping track of events, activities, interactions that occur during database system operation.
Retrieved from "https://www.infra-repository.org/oiar/index.php?title=PD.Relational_Data_Management&oldid=1586"