Hi, I'm looking into setting up an on-prem Lakehouse with an S3 compatible object store, and k8s compute. I'd like to be able to use multiple processing engines (Spark, Flink, Presto...) I'm thinking about using Delta Lake as the table format, but I'm having trouble finding a Metadata Catalog that supports Delta Lake. (Do I even need it ?) If I were to choose Iceberg, I would use Nessie. Any suggestions would be appreciated. Thanks!