NetApp StorageGRID, like AWS S3, adheres to standards for decoupling data from metadata. However, the product faces challenges in this area, particularly concerning the storage of metadata. Unlike some other solutions, StorageGRID stores all metadata in the first storage node, and expanding metadata capacity necessitates adding more storage nodes. This design choice, using Cassandra database for metadata storage, can lead to significant costs as organizations may need to add more storage nodes solely to accommodate metadata requirements, even if additional storage capacity isn't needed. The challenge lies in the taxing aspect of managing both small and large objects efficiently. On average, the metadata space for objects ranges from 1.3 to 2 KB, regardless of object size. While this variability is common across platforms, it poses specific challenges within StorageGRID. Managing small object transactional records stored as objects creates pain points within storage groups. One key improvement I'd like to see in StorageGRID is enhanced visibility for management purposes. For instance, tracking delete markers, which is a standard mechanism, becomes challenging when bucket versioning is enabled. Currently, there's a lack of visibility in this area, unlike what is offered by AWS. However, I'm aware that product enhancements are underway to address this limitation. While StorageGRID's documentation is comprehensive in some aspects, it could be stronger in providing details on certain events and metrics.