I’m copying objects from a Standard bucket to a bucket intended for S3 Glacier Deep Archive, but some destination objects show Glacier Flexible Retrieval. What settings or bucket behavior should I check?
Why is DataSync putting some objects in Glacier Flexible Retrieval?
Asked by
Sprocket-6Q
Answers
Check the storage class configured on the DataSync destination location. New objects use the class specified there, so make sure it is set to Deep Archive rather than relying on the destination bucket’s name or intended use.
Reply to this answer
Also look at the task’s handling of objects that already exist at the destination. DataSync can preserve their existing storage class unless the task is configured to transfer all data. That could explain why only some objects differ.
Reply to this answer
Inspect the destination bucket’s lifecycle rules too. A rule can transition objects to another storage class after they arrive, so compare the object timestamps and current class with the rule configuration.
Reply to this answer
If the goal is just to archive objects already in S3, it may be worth checking whether DataSync is needed for this workflow. The right method depends on why you are copying them, but the storage-class setting still needs to be explicit.
Reply to this answer
Add your answer
Answers are reviewed before publication.
Sprocket-6Q:
So I should check both the DataSync location and any lifecycle transitions, rather than assuming the bucket’s target class controls every copy?