Enable Active-Passive Bucket Replication

The procedure on this page creates a new bucket replication rule for one-way synchronization of objects from one AIStor bucket to another AIStor bucket. The buckets can be on the same AIStor deployment or on separate AIStor deployments.

Replication requires all participating clusters meet the bucket replication requirements. This procedure assumes you have reviewed and validated those requirements.

1. Log in to the AIStor Console for the deployment.

2. Select the Manage button for the bucket to replicate.

3. Select the Replication section.

4. Select Add Replication Rule +.

5. Complete the requested information:

   Field	Description
   Priority	Enter a number value to indicate the order in which to process replication rules for the bucket. 1 indicates the highest importance.
   Target URL	The URL of the deployment to replicate data to.
   Use TLS	Leave the toggle in the ON position if the destination deployment uses TLS. Otherwise, move the toggle to the OFF position.
   Access Key	The user name to use on the destination deployment. The user must have write access to the bucket to replicate to.
   Secret Key	The password for the provided Access Key.
   Target Bucket	The bucket at the destination to write the data to. The target bucket may have the same name as the origin bucket, depending on the destination bucket location.
   Region	The AWS resource region location of the destination deployment.
   Replication mode	Leave the default selection of Asynchronous to allow AIStor to replicate data after the write operation completes on the origin ment. Select Synchronous to attempt to complete the replication of the object during its write operation. While synchronous replication may result in more reliable synchronization between the origin and destination buckets, it may also increase the time of each write operation.
   Bandwidth	Specify the maximum amount of bandwidth the replication process can use while replicating data. Enter a number and select a data unit.
   Health Check Duration	The maximum length of time in seconds AIStor should spend verifying the health of the replicated data on the destination bucket.
   Storage Class	The class of storage to use on the destination deployment for the replicated data. Valid values are either STANDARD or REDUCED_REDUNDANCY.
   Object Filters	Limit which objects to replicate from the bucket by prefix or tags. If you enter multiple tags, the objects must match all tag values.
   Metadata Sync	Leave selected to also replicate the object’s metadata file. Otherwise, move the toggle to the Off position.
   Delete Markers	Leave selected to also replicate AIStor’s indication that an object has been deleted and should also be marked deleted at the action bucket. Otherwise, move the toggle to the Off position to prevent marking the object as deleted in the destination bucket.
   Deletes	Leave selected to allow replication of the deletion of versions of an object. Otherwise, move the toggle to the Off position to not replicate deletion of object versions.

6. Select Save to finish adding the replication rule.

To validate the configuration, upload an object to the bucket on the source deployment. Then check the bucket on the remote deployment and ensure the object replicates successfully.

1. Add the new replication rule to the source AIStor deployment.

   Use the mc replicate add command to add a new replication rule to the source AIStor deployment.

   mc replicate add ALIAS/BUCKET \
      --remote-bucket 'https://USER:PASSWORD@HOSTNAME:PORT/BUCKET' \
      --replicate "delete,delete-marker,existing-objects"
   

   • Specify the ALIAS and Bucket of the origin AIStor deployment.
   • Modify the --remote-bucket parameters to specify the remote AIStor deployment and bucket

   Once saved the replication rule synchronizes data from the ALIAS/BUCKET to the remote BUCKET.

2. Validate the configuration

   Use mc cp to copy a new object to the replicated bucket on one of the deployments.

   mc cp ~/foo.txt ALIAS/BUCKET
   

   Use mc ls to verify the object exists on the destination bucket:

   mc ls ALIAS/BUCKET