Amazon S3 has been used in a wide variety of ways. Customers have given feedback regarding Amazon’s service, and this is how the idea of cross-region replication was born. S3 objects copies can be made easily in another AWS region with the help of Cross region replication. This feature has been built over S3’s existing versioning facility. It can be used to copy files which are present in geographically diverse locations.
When the cross-region replication is enabled, an object which is uploaded to a specific S3 bucket gets automatically duplicated into a destination bucket that is located at a different AWS location.
The objects in the source and destination buckets are copied in an asynchronous manner. For this object replication, bucket-level configuration is necessary. The replication configuration has to be added to the source bucket.
The minimum requirements include:
- A destination S3 bucket where objects are replicated to.
- An AWS IAM role which S3 considers replicating objects on behalf of the user.
Features of cross-region replication
- It helps seamlessly copy object from an S3 bucket in a specific location to another location’s S3 bucket.
- It can be implemented only when both the buckets have their versioning feature enabled.
- When an object is being copied from one bucket to another, data is encrypted with the help of SSL so as to protect it from attacks.
- If data which is being copied from source bucket is already present in the destination bucket, data doesn’t get replicated even if the operation is performed. This way, redundant data is not stored.
When should cross region replication be used?
It can be used in the following cases:
- Following the compliance requirements: When the compliance requirements need to be met: Amazon S3 allows the user to store data across various regions all over the world in specific Availability Zones by default, but compliance rules specify the exact region where data has to be stored (at much longer distances).
- Reducing the latency: If two users are in two different locations of the world, objects required by them can be accessed much more quickly by storing copies of the objects closer to the respective user’s locations.
- When two owners need to maintain copies of objects: When the same set of objects need to be used to analyse data, and two compute clusters require this data, wherein the clusters are present in different AWS Regions, object copies can be maintained in both the regions to improve the efficiency.
In this post, we understood one of the newly introduced features of AWS, the cross-region replication which helps copy object data from one S3 bucket present in one AWS region into another bucket, which is present in a different region seamlessly.