OGG architecture diagram / pattern / category

Note: This article is an overview of the Oracle GoldenGate architecture / OGG architecture diagram / OGG mode introduction / OGG mode classification.
Reminder: If you find that there is a problem in this article or there is a better way to write it, please leave a message or private letter to modify and optimize

★ Introduction

Capture process: Oracle GoldenGate capture process is called Extract. Each instance of the extraction process is called a group, which includes the process itself and related documents that support it.
Transfer process: It is recommended to use another extraction process (called data pump) on the source system in order to persist the captured data locally to a series of files (called trace). The data pump does not capture the data, but reads the local trace and propagates the data to the target through the network.
Application process: The Oracle GoldenGate application process is called Replicat. Each instance of a Replicat process is called a group, which includes the process itself and related files that support it. Replicat reads the data sent to local storage (called tracking) and applies it to the target database.

★ Overall architecture diagram

★ Mode details

Group type	Mode name	brief introduction
extract	classic	In classic capture mode, the Oracle GoldenGate Extract process captures data changes from the Oracle redo or archive log files on the source system or the archive logs attached to the standby system
extract	integrated (integrated mode)	In the integrated capture mode, the Oracle GoldenGate Extract process directly interacts with the database log mining server to receive data changes in the form of logical change records (LCR)
replicate	integrated (integrated mode)	In integrated mode, the replication process takes advantage of the application processing functions available in the Oracle database.
	nonintegrated (non-integrated mode)	In non-integrated mode, the replication process is a single-threaded process that uses standard SQL to apply data to the target table
	coordinated (collaborative mode)	The difference between classic mode and collaboration mode is that replication is multi-threaded in collaboration mode. In a single Replicat instance, multiple threads independently read traces and apply transactions in parallel. Each thread handles filtering, mapping, conversion, SQL construction, and error handling for its assigned workload. A coordination thread coordinates the dependencies between threads of a cross-thread trading account.