Databases 20 min read

Mastering Oracle GoldenGate: Architecture, Components, and Configuration Guide

This article provides a comprehensive overview of Oracle GoldenGate, detailing its supported databases, modular architecture, key components such as Extract, Data Pump, Replicat, Trails, Checkpoints, Manager and Collector, as well as processing types, group configuration, and commit sequence numbers for reliable data replication.

ITPUB
ITPUB
ITPUB
Mastering Oracle GoldenGate: Architecture, Components, and Configuration Guide

Oracle GoldenGate enables real‑time data exchange and manipulation across heterogeneous IT platforms at the transaction level, ensuring data integrity with minimal overhead. It supports a wide range of databases and DDL operations, offering flexible extraction, filtering, and custom processing to meet diverse business needs.

1. Supported Databases and Processing Methods

The tool works on various databases, handling both initial data loads and continuous change synchronization while preserving transaction consistency. Detailed installation and configuration guidance is available in the official Oracle GoldenGate documentation.

2. Architecture Overview

The logical architecture consists of several components:

Extract (data extraction)

Data Pump (data transport)

Replicat (data apply)

Trails or extract files (temporary storage)

Checkpoints (position tracking)

Manager (process control)

Collector (target‑side helper)

GoldenGate logical architecture
GoldenGate logical architecture

2.1 Extract Overview

The Extract process runs on the source or downstream database (or both) and captures data at the source. It can be configured for:

Initial loads : static extraction of current data sets.

Change synchronization : continuous capture of DML and DDL changes.

Data sources include source tables, database recovery/transaction logs, and third‑party capture modules. Extract writes captured operations to trail files, preserving order until a commit or rollback is received.

2.2 Data Pump Overview

Data Pump reads local trail files and transports data to remote trail files over the network, adding flexibility and isolation from TCP/IP activity. It supports filtering, mapping, transformation, and a pass‑through mode that bypasses data manipulation for higher throughput.

Protection against network or target failures.

Multi‑phase data filtering and transformation.

Consolidation of data from many sources.

Synchronization of one source to multiple targets.

2.3 Replicat Overview

Replicat runs on the target database, reads trail files, and applies captured DML/DDL using dynamic SQL and bind variables. Configuration options include:

Initial loads : bulk loading of a data backup set.

Change synchronization : applying captured changes via local DB interfaces or ODBC.

Multiple Replicat processes can run in parallel, either in a coordinated (threaded) mode or an integrated mode that leverages native Oracle processes. Delayed apply is controlled by the DEFERAPPLYINTERVAL parameter, while performance‑related parameters such as BATCHSQL, GROUPTRANSOPS, and MAXTRANSOPS are documented in the Oracle GoldenGate reference.

2.4 Trails Overview

Trails are sequential files that temporarily store captured changes. They can reside on the source, an intermediate system, or the target. Local trails are called extract files; remote trails are called remote trails. Trails enable decoupled extraction and replication, allowing independent processing and flexible data movement.

2.5 Checkpoints Overview

Checkpoints record the current read/write position of processes, enabling recovery after failures. Extract creates checkpoints for source logs and trail positions; Replicat writes checkpoints to a checkpoint table in the target database and to checkpoint files. The checkpoint interval is controlled by the CHECKPOINTSECS parameter.

2.6 Manager Overview

The Manager process must run before Extract and Replicat, handling process startup, dynamic process control, port management, trace management, and event/error reporting. One Manager can control multiple Extract or Replicat processes.

2.7 Collector Overview

Collector runs on the target side, receiving connection requests from remote Extract processes, binding ports, and writing received changes to trail files. It can be initiated by the Manager or, in some configurations, by the target side.

3. Processing Types

GoldenGate supports several processing modes:

Online Extract/Replicat with checkpoints for continuous, recoverable replication.

Source‑is‑table (initial load) Extract without checkpoints.

Special‑run Replicat for known start/end positions, useful for batch loads.

Remote task (RMTTASK) where Extract communicates directly with Replicat over TCP/IP, bypassing Collector and trail files.

4. Groups Overview

Process groups distinguish multiple Extract and Replicat processes within the same environment, allowing parallel replication of different data sets. Groups share parameter files, checkpoint files, and, for Replicat, a checkpoint table. Groups are defined with ADD EXTRACT and ADD REPLICAT commands.

5. Commit Sequence Number (CSN) Overview

The CSN uniquely identifies the commit point of a transaction, ensuring consistency and data integrity across source and target. It can be used by Extract to locate the correct log position and by Replicat to position trails accurately. Detailed CSN values per database are listed in the appendix of the Oracle GoldenGate documentation.

GoldenGate architecture diagram
GoldenGate architecture diagram

To upgrade trail file sequence length from 6 to 9 characters, stop the Extract process and run: convchk extract trail seqlen_9d To downgrade back to 6 characters, use: convchk extract trail seqlen_6d Trail files are created automatically; their names consist of a two‑character prefix followed by a 6‑ or 9‑digit sequence (e.g., tr000001). When the sequence reaches its maximum, it rolls over. Old trails can be purged with the PURGEOLDEXTRACTS parameter.

Original Source

Signed-in readers can open the original source through BestHub's protected redirect.

Sign in to view source
Republication Notice

This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactadmin@besthub.devand we will review it promptly.

data replicationETLdatabase migrationChange Data CaptureOracle GoldenGate
ITPUB
Written by

ITPUB

Official ITPUB account sharing technical insights, community news, and exciting events.

0 followers
Reader feedback

How this landed with the community

Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.