Suspiciously small snapshot size for replicated Vault cluster (snapshot-auto)

Hello, I have 2 different Vault clusters serving different environments. Environment A (single vault cluster) produces snapshots around ~500mb. This was not a surprise.

Environment B (the replicated cluster) produces snapshots around ~2mb. Is this indicative of some type of error? I have no immediate way to test a restore to find out if it is corrupted or there is some type of error.

Also, the documentation is not very helpful here:

Each Vault cluster (forming the Primary and Secondary Vault Performance Replicas) should be configured to create separate automated snapshots.

How exactly would one do this?

Here is the log for the snapshot creation in Environment B:

[DEBUG] core.snapshotmgr.s3_backup: snapshot complete: name=s3_backup elapsed=414.506664ms size=0

size=0 looks wrong to me, but I do not know how to troubleshoot further.

What is the replicated cluster? Is it a DR or PR?

What’s the backend for each?

Sorry should have included that info. It is PR, raft backend storage.

Okay. Assuming it isn’t sealed, check your token policy, sounds like it doesn’t have the rights to do the snapshot.