Prysm unable to provide expired blob even when configured to

[parithosh]

Describe the bug

A prysm node was setup pre-dencun mainnet fork with the flag "--blob-retention-epochs=300000000",. This should in theory configure the node to store blobs beyond the expiry window.

The API call curl localhost:5052/eth/v1/beacon/blob_sidecars/8626178 queries the first slot with a blob. This call worked fine inside the blob expiry window.

The same call now returns {"data":[]}, indicated the blob can't be read via the API beyond the expiry window. The call was also checked against a lighthouse node with a similar configuration and it indeed returns the blob as expected.

The DB size however hasn't reduced and is continuing to increase:
37G mainnet_archive_blob_storage_prysm_nethermind_001, so its indicative of the blob being present but unable to be retrieved via the API.

Has this worked before in a previous version?

n/a

🔬 Minimal Reproduction

Query a node that has been configured to store blobs forever (theoretically)

Error

n/a

Platform(s)

Linux (x86)

What version of Prysm are you running? (Which release)

v5.0.3

Anything else relevant (validator index / public key)?

No response