Oracle database (online) replication

Question

I am a Java guy, I can work with Oracle Database, I know PLSQL, SQL. But I am not good at managing database servers. I think it is a completely different area.

My question is related to database replication. I googled it, found millions of answers but I am still confused.

I could see many times in my professional carrier that developers create complete (complicated) applications to keep sync the source database schema to a target one. It takes time to develop sync apps and very hard to maintain them, especially in case of any data structure modification for example in tables.

I could see that apps built with JPA, JDBC, Spring, myBatis, and PLSQL as well. Usually, they sync DBs during the night, scheduled by Cron, Quartz, Spring, etc. During the sync process usually, the source DB is only available for querying data, not for inserting and DB constraints and triggers are disabled.

These kinds of custom applications always scare me. I do not believe that there is no general, easy, and official way to keep sync two databases without developing a new application.

Now, I got a similar task and honestly, I would like to write zero lines of code related to this task. I believe that there are recommended and existing solutions, cover this topic offered by the database vendors.

That would be great if you can push me in the right direction. I feel that writing another new DB sync application is not the right way.

I need to focus on Oracle Database sync, but I would be happy to know a general, database vendor-independent way.

Answer 1

There are many ways to perform replication in a Oracle Database. Oracle has two replication techniques in the database "Advanced Replication" and "GoldenGate". GoldenGate us the new perferred method of replication which uses the redo logs files from the database. Both methods are geared for a Oracle DBA.

Often application developers will create a "interface" that will move data from one database to other. A interface is a program ( pl/sql, bash, c, etc ) that runs on a cron (database or system) that wakes on a event to move data. Interfaces are useful when data is needed to be process during replication.