Amazon Relational Database Service (RDS), now supports launch read replica in another region. It takes just a few clicks. THIS IS HUGE! This is the first feature that AWS supports across all regions. This is AWS’s first major step towards better support for infrastructure setup across regions.
Replicas were previously restricted to the same Region. RDS instances do not allow access to the server but only the database application. Until now, it was impossible for RDS to configure and read replicas in other regions. AWS users can now, with just a few clicks start reading replicas in another area.
This feature is very important for many uses:
Disaster recovery
Although it is rare for an entire AWS region or network to go down, it can happen. Many companies want to create multiple copies of their databases across regions so that infrastructure can be quickly set up in another region in the event of a catastrophe. This setup requires that databases are synced across multiple regions. Until now, RDS was not available for such deployments. They can now.
Active-Active cross-region deployments
This was a case for active-active deployments that span over regions. An e-commerce site might need infrastructure in the USA and the UK simultaneously to service clients locally. However, the product database could have been identical. This requires database sync across countries, and RDS was not an option. At the very minimum, both locations will have local read-replica copies that can be used to read data. However, the writes must still be sent to a central location. This arrangement would be ideal for heavy read-intensive applications such as e-commerce.
Data migration
Sometimes, even though the applications and databases may be located in different regions, it might be necessary to periodically copy data to one region for all other regions. This could be used for data aggregation in a central region for analysis. Imagine a large IT company with infrastructure in all regions. However, daily reports that require all the data are generated using aggregation data from all regions. For such cases, RDS was not an option. Now, there are multiple databases that could have their own RDS masters and a read replica. This allows data from all regions to be processed locally.
This setup can be used for many other purposes. What other use cases do this feature allow? Please share your thoughts! Please share this article if you enjoyed it.
