Managing HBase databases

ONLINE HELP
WINDEV, WEBDEV AND WINDEV MOBILE

Version:

Home | Sign in | English

This content has been translated automatically. Click here to view the French version.

Help / WLanguage / Managing databases / Big Data / Managing HBase databases

Overview
How is an HBase database organized?
How to handle an HBase database in programming?
How to create an HBase table?
How to retrieve the list of tables found in an HBase database?
How to retrieve the description of an HBase database?
How to delete an HBase table?
How to read the data found in an HBase database?
How to write data into an HBase database?
How to delete data from an HBase table?
How to perform a browse with filter in an HBase table?

WINDEV

WEBDEV

WINDEV Mobile

Others

See also

HBase functions

Managing HBase databases

Overview

HBase is a non-relational distributed database (NoSQL), based on columns and using a non-structured storage for large tables.

HBase is a core component of Hadoop.

HBase has been conceived for processing large volumes of data (several million rows, several million columns per row) and non-standard data (each row can have a structure different from the next row).

How is an HBase database organized?

In an HBase database, the data is stored in tables (files), in a denormalized way.
Each table is divided into families of columns (items), that themselves contain columns.
Each row (equivalent to a record in a traditional DBMS) is identified by a single "rowKey".
For each row, in the same family of columns, the number and the name of columns is free.

The "rowKey" is the single key of the table.
A cell is the intersection between a row and a column.
The history of the cell content is stored as well as its timestamp (date/time when it was written).
You have the ability to implement filters on an HBase table (a little bit like a filter on HFSQL).

How to handle an HBase database in programming?

WLanguage allows you to:

create tables
use existing tables:
use the table data:

Please note: HBase functions are available for editing and execution only on Windows Vista or Windows Server 2008.

How to create an HBase table?

To create an HBase table, you must:

Define and initialize a hbConnection variable.
Define and initialize a hbTableDescription variable. This variable contains an array of hbColumnDescription variables describing the columns.
Use the hbCreateTable function.Example:

// Connection
Connection is hbConnection
Connection.Server = "MyServer"
// Describe the table
TableDesc is hbTableDescription
TableDesc.Name = "blog"
// Describe the "Post" column
ColDesc is hbColumnDescription
ColDesc.Name = "Post"
Add(TableDesc.Column, ColDesc)
// Create the table
hbCreateTable(Connection, TableDesc)

How to retrieve the list of tables found in an HBase database?

To retrieve the list of tables found in an HBase database, you must:

Define and initialize a hbConnection variable.
Use the hbListTable function.Example:

// Connection
Connection is hbConnection
Connection.Server = "MyServer"
// Display the list of tables
Trace(hbListTable(Connection))

How to retrieve the description of an HBase database?

To retrieve the description of an HBase database, you must:

Define and initialize a hbConnection variable.
Use the hbGetTableDescription function.
Retrieve the table description in a hbTableDescription variable. This variable contains an array of hbColumnDescription variables describing the columns.Example:

// Connexion
Connexion is hbConnection
Connexion.Server = "MonServeur"
// Récupération de la description de la table
DescTable is hbTableDescription = hbRécupèreDescriptionTable(connection, "blog")
// Affichage du nom de la table
Trace("table: " + DescTable.Name)
// Affichage de la description des colonnes
FOR EACH DescCol OF DescTable.Column
	Trace(TAB + "colonne: " + DescCol.Nom)
	Trace(TAB + TAB + "attribut: ")
	FOR EACH ValAttribut, NomAttribut of DescCol.Attribut
		Trace(TAB + TAB + NomAttribut + " = " + ValAttribut)
	END
END

How to delete an HBase table?

To delete an HBase table, you must:

Define and initialize a hbConnection variable.
Use hbDeleteTable and specify the name of the table to delete.Example:

// Connection
Connection is hbConnection
Connection.Server = "MyServer"
// Delete the table
hbDeleteTable(Connection, "blog")

How to read the data found in an HBase database?

To read the data found in an HBase database, you must:

Define and initialize a hbConnection variable.
Define and initialize a hbReading variable. This variable contains an array of hbColumn variables describing the columns to retrieve.
Use the hbRead function.
Read data in the hbCell resultExample:

// Connexion
Connexion is hbConnection
Connexion.Serveur = "MonServeur"
// Description des données à lire
Lecture is hbReading
Lecture.Ligne = "post1"
Lecture.Version = 2
// Description colonne
Col is hbColumn
Col.Famille = "Post"
Add(Lecture.Colonne, Col)
// Lecture des données
MonRésultat is array of hbCell = hbLit(connection, "blog", Lecture)
// Affichage des données lues
FOR EACH Cellule OF MonRésultat DO
	Trace("Colonne: " + Cellule.Colonne.Qualificatif + "; Valeur: " + ...
	Cellule.Valeur + "; timestamp: " + Cellule.Timestamp)
END

How to write data into an HBase database?

To write data into an HBase database, you must:

Define and initialize a hbConnection variable.
Define and initialize a hbWriting variable. This variable contains an array of hbCell variables describing the data to write.
Use the hbWrite function.Example:

// Connection
Connection is hbConnection
Connection.Server = "MyServer"
// Description of data to write
Writing is hbWriting
Writing.Row = "post1"
Cell is hbCell
Cell.Column.Family = "post"
Cell.Column.Qualifier = "title"
Cell.Value = "This is my blog post."
Add(Writing.Cell, Cell)
// Write data
hbWrite(Connection, "blog", Writing)

How to delete data from an HBase table?

To delete data from an HBase table, you must:

Define and initialize a hbConnection variable.
Use hbDelete and specify the identifier of the row to delete.Example:

// Connection
Connection is hbConnection
Connection.Server = "MyServer"
// Delete the entire row
hbDelete(Connection, "blog", "post1")

How to perform a browse with filter in an HBase table?

To perform a browse with filter in an HBase database, you must:

Define and initialize a hbConnection variable.
Define and initialize a hbScanParameter variable. This variable defines the filter to implement.
Browse the filtered data (corresponding to the hbScanParameter variable) via the FOR EACH statement. The browse result is read in a hbScanResult variable. Example:

// Connexion
Connexion is hbConnection
Connexion.Server = "MonServeur"
// Description du filtre
MonScan is hbScanParameter
MonScan.Connection = connection
MonScan.Table = "blog"
// Filtre pour toutes les lignes les valeurs des colonnes dont le nom commence par "Titre"
MonScan.Filter = "{ ""type"": ""ColumnPrefixFilter"", ""value"": """ + ...
	Encode("Titre", encodeBASE64) + """ }"
// Parcours selon le filtre et affichage des données filtrées
// La variable de parcours MonRésultat est de type hbRésultatScan
FOR EACH MonRésultat OF MonScan
	Trace("La colonne " + MonRésultat.Colonne.Famille + ":" + ...
		MonRésultat.Colonne.Qualificatif + ...
		" de la ligne " + MonRésultat.Ligne + " vaut " + MonRésultat.Valeur)
END