Replication

Last updated:2021-09-14 16:17:51

Cross-region replication automatically replicates objects across KS3 buckets in different regions. It synchronizes object operations, including creation, overwriting, and deletion, from a source bucket to a destination bucket in a different region. Objects in the destination bucket are the exact replicas of the objects in the source bucket. They have the same object names, metadata, and content as the source objects, such as the same creation time, owner, user-defined metadata, object ACL, and object content. Cross-region replication allows you to replicate unencrypted objects and objects encrypted on the server side.

Notes:

  • After cross-region replication is enabled, you are charged for the traffic generated when objects are replicated from the source bucket to the destination bucket in a different region. Each time an object is synchronized, KS3 accumulates the number of requests and charges you on a Pay-As-You-Go basis. If the source object is of the IA storage class, KS3 also charges you a retrieval fee.
  • Traffic fees of cross-region replication are charged on a Pay-As-You-Go basis. KS3 does not provide resource packages. For more information, see Billing modes and items.
  • Cross-region replication is applicable to regions in mainland China and Hong Kong.
  • After cross-region replication is enabled, it takes effect within 1 to 6 minutes, and then data is replicated in near real time.
  • Cross-region replication is applicable only to the data generated after replication is enabled, and historical data is not replicated.

Scenarios

  • Security compliance: Although KS3 keeps multiple replicas of each stored object on physical disks by default, remote disaster recovery is required for the specified data according to the security compliance requirements. Cross-region replication allows you to replicate data across remote data centers to meet the security requirements.
  • Latency reduction: To minimize the latency when users access objects, you can replicate objects between two data centers that are geographically close to users.
  • Data backup and disaster recovery: Cross-region replication ensures high data security and availability. If one data center fails due to an earthquake, a tsunami, or other disasters, backup data in another data center can be used to restore business as soon as possible.
  • Data replication: If you need to migrate data from one data center to another or use different replicas of different data centers for business reasons, replication can meet your requirements.

Operation methods

Operation method Reference Description
Console Configure cross-region replication Web application, which is easy to use.
SDK Use the SDK for Java to perform cross-region replication Demo of the SDK for Java

Usage notes

If you have two buckets in different regions, you can configure synchronization rules to replicate data from the source bucket to the destination bucket in real time. Cross-region replication has the following features:

  • Real-time data synchronization: Data is replicated in real time. Data additions, deletions, and modifications can be synchronized to the destination bucket in real time. A 2-MB object can be replicated within minutes to ensure the eventual consistency of data in both buckets.
  • Two-way synchronization: You can configure replication from Bucket A to Bucket B, and replication from Bucket B to Bucket A, to achieve two-way replication.
  • Transfer: You can configure replication from Bucket A to Bucket B, and replication from Bucket B to Bucket C.
  • Encrypted replication: You can replicate unencrypted objects and objects encrypted on the server side.

Limits

  • For two synchronized buckets, objects replicated from the source bucket may overwrite the objects with the same names in the destination bucket.
  • In cross-region replication, data is replicated asynchronously in near real time. Based on the object size, it takes minutes to hours to replicate data from the source bucket to the destination bucket.
  • Three replication modes are supported: one-way, two-way, and transfer. A bucket cannot serve as a source bucket or destination bucket for multiple replication relationships. For example, data in Bucket A cannot be replicated to both Bucket B and Bucket C, and Bucket A cannot serve as a destination bucket for both Bucket B and Bucket C.
  • Multiple buckets with cross-region replication enabled must belong to different regions, and data cannot be replicated between buckets in the same region.
  • Data is replicated only once. If you configure replication from Bucket A to Bucket B and replication from Bucket B to Bucket C, an object that is replicated from Bucket A to Bucket B will not be further replicated to Bucket C.
  • If you modify the object in the source bucket before the replication is completed, the modification may not be replicated to the destination bucket.

Did you find the above information helpful?

Unhelpful
Mostly Unhelpful
A little helpful
Helpful
Very helpful

What might be the problems?

Insufficient
Outdated
Unclear or awkward
Redundant or clumsy
Lack of context for the complex system or functionality

More suggestions

0/200

Please give us your feedback.

Submitted

Thank you for your feedback.

问题反馈