Why does S3 Inventory service create "inventory/data/" key?

Question

We have an S3 bucket that uses S3 Inventory to store records of all keys in the bucket, and the inventory files are stored on the bucket itself (as opposed to a different bucket) under the prefix `inventory/` — i.e., manifests are stored at keys like `inventory/2024-12-19T01-00Z/manifest.json`, and the inventory list files are stored at keys like `inventory/data/$UUID.csv.gz`.

We recently discovered that there exists an `inventory/data/` key (with a trailing slash) on our bucket, presumably created by the S3 Inventory service.  This abnormal key creates problems for our backup process.  Why did S3 Inventory create this key?  Can we safely get rid of it?

Leo K · Answer

Empty objects with keys ending in forward slashes are used by tools like the S3 management console to display a conceptual "folder" while it doesn't "contain" any objects (and more accurately, when no current object versions exist with keys matching that "folder" prefix). S3 inventory does seem to create the data/ object for this purpose also in my environment. There shouldn't be any adverse effects if you just delete the empty "folder placeholder," but I can't say for sure that it couldn't reappear in some situation, such as when the inventory configuration is updated.

EXPERT

Leo K

answered 2 years ago

EXPERT

Riku_Kobayashi

reviewed 2 years ago

Why does S3 Inventory service create "inventory/data/" key?

Add your answer

Relevant content