How MSDP S3 object import works
The following steps describe how MSDP S3 object import synchronizes objects between the production and replica domains.
In the production domain, an MSDP-Object-Store backup policy runs and NetBackup creates a backup image of the S3 bucket's objects.
NetBackup starts a replication job, through the configured AIR SLP, to replicate the image to the replica domain.
When the replication job finishes, the production MSDP storage server generates an additional import event in the replica domain, alongside the standard AIR image import event. This event carries the information NetBackup needs to import the replicated image into the replica MSDP S3 catalog.
In the replica domain, NetBackup polls for and picks up the import S3 event, and creates an import into S3 bucket job. This job calls the MSDP S3 server to start importing the object metadata.
The MSDP S3 server reads the object listing captured in the replicated backup image and loads that metadata into the S3 service, making the objects available in the replica bucket.
NetBackup periodically polls the MSDP S3 server until the import completes, and updates the job status in the Activity Monitor.
This same sequence can run in the opposite direction, importing objects from the replica domain back into the production domain's bucket, provided that S3 import is also enabled for the corresponding policy on the replica MSDP storage server.
