We're building a data lake to serve as our central data hub - primarily to support AI workloads and model training, but also to power both internal and customer-facing analytics. We chose OCI mainly for cost reasons: while the raw storage costs aren't drastically different from AWS, the egress and compute costs add up quickly.

We're already a Fivetran customer for other projects, but I don't think we have any flexibility on our storage backend for this data set...which unfortunately means Fivetran is likely out of the running for this one. It looks like HVR does support OCI for destination storage, but its feature set obviously doesn't provide the same value as the managed data lake service.

As another feature request (happy to start a new topic if preferable) it would also be great if you supported writing Hive catalogs - and Hadoop-style catalogs would also be nice. Hive metastore support is essential for anyone aiming to maintain provider-agnostic flexibility. Hadoop-style catalogs are a nice-to-have just for the simplicity and portability they provide, especially for smaller or simpler use cases where an external service dependency would be overkill.