architecture du serveur oracle

Introduction

  • Overview of the Oracle Server Architecture

  • Importance of understanding database administration

Architecture du serveur de base de données Oracle

1. Serveur de base de données Oracle

  • Integration of Oracle instance and database files.

2. Composantes du serveur Oracle

  • Instance Oracle:

    • Structure composed of a shared memory area called SGA (System Global Area).

    • Includes background processes with specific roles.

  • Base de données Oracle:

    • Consists of data files, control files, and redo log files.

3. Démarrage de la base de données Oracle

  • Fichiers de paramètres d’initialisation:

    • Critical for configuring database parameters at startup.

4. Arrêt de la base de données Oracle

  • Procedures for shutting down the database safely.

Démarrage de la base de données Oracle

Scénario de connexion à une base de données Oracle

  1. Client contacts the Oracle listener selecting instance.

  2. Listener initiates a dedicated server process.

  3. Listener sends acknowledgment to client with server process address.

  4. Client establishes connection with dedicated server process.

  5. Server process connects to Oracle instance on behalf of the user.

  6. User processes do not interact directly with the Oracle server.

Instance Oracle

SGA (System Global Area)

  • Components of SGA:

    • Mandatory areas:

      • Shared Pool

      • Database Buffer Cache

      • Redo Log Buffer

    • Optional areas:

      • Large Pool

      • Java Pool

      • Streams Pool

Shared Pool

  • Cache Components:

    • Cache du dictionnaire de données:

      • Stores user account data, tables, index descriptions.

      • Enhances performance by reducing response time for database instructions.

    • Library Cache:

      • Stores SQL command and PL/SQL code recently used.

      • Optimizes resource use by reusing existing commands.

Database Buffer Cache

  • Stores data blocks extracted from data files.

  • Managed by an LRU (Least Recently Used) algorithm.

Tampon de journalisation

  • Records all modifications made to data blocks.

  • Essential for data recovery and managed by LOG BUFFER parameter.

Processus en arrière-plan

DBWn (Database Writer)

  • Writes modified blocks from database buffer cache to data files.

  • Triggers on various conditions: memory limits, timeouts, etc.

LGWR (Log Writer)

  • Writes log entries to disk for recovery purposes.

  • Triggers at specific memory usage thresholds or time intervals.

SMON (System Monitor)

  • Responsible for instance recovery.

  • Manages space merging and temporary segment releases.

PMON (Process Monitor)

  • Cleans up user processes after failures.

  • Frees locks and system resources.

CKPT (Check Point)

  • Coordinates with DBWn to mark recovery points.

  • Ensures up-to-date file headers and control files.

PGA (Program Global Area)

  • A memory structure created for each connected user.

  • Stores session-specific control information and is released upon disconnection.

Conclusion

  • Understanding Oracle's architecture is critical for database administration tasks.