目次
Hi, I'm an AWS engineer.
@Darcy
I am an AWS engineer.
In this issue, we would like to learn about cross-region replication of S3.
As anyone who has considered backing up storage will know, it is impractical to copy the entire large storage capacity all at once.
Cross-region replication of S3 would be a solution to answer such questions.
S3] About cross-region replication
Between two buckets in different AWS regions
between two buckets in different AWS regions.
objects between two buckets in different AWS regions.
The functionality of this feature is to automatically copy objects asynchronously between two buckets in
Replication is triggered by data processing events such as object creation, update, and deletion for a bucket.
Cross-region replication can be set up for buckets owned by the same AWS account or by different accounts.
What are we doing it for?
use case
use case
Web applications are used for multiple EC2 instances.
When accessing S3 in another region from this EC2 instance, the
VPC endpoints.
It appears that the current architecture does not allow VPC endpoints to be used as is.
For example, when using a separate VPC endpoint to share content in an S3 bucket.
feature
Cross-region replication can be set up for buckets owned by the same AWS account or by different accounts.
Cross-region replication is performed simultaneously with object registration and deletion
Can be executed manually by AWS CLI command operation
On the management console, objects are automatically replicated once they are registered.
It is possible to configure settings to enable cross-region replication on the management console, but it is not possible to perform operations to synchronize between buckets at this stage.
Bidirectional replication is possible, but requires separate bidirectional cross-region replication setup.
VPC endpoints are points within a region, so they are not accessible from outside the region.
Therefore, to use the endpoints, S3 objects must be copied by S3 cross-region replication to the region where they will be used.
This keeps the S3 objects replicated to the target region and configures access by VPC endpoints to that region.
It can be executed manually by command operation of the AWS CLI.
way
Replicate operations on the management console
On the management console, replication is automatic once the object is registered.
It is possible to configure the management console to enable cross-region replication, but not the operation to synchronize between buckets at this stage.
In this case, use the AWS CLI command.
Always monitor for redundancy
Once constructed, the following operational procedures must be considered
Which region stores which named object, in which size
When is it synchronized?
Are they the same size?
How do I know if it's the right move?
There are many ways to conduct surveillance.
