The iceberg documentation has a section on maintenance for streaming tables which may give you some ideas you can try (more details in the linked documentation):

• Expire old snapshots

• Compacting data files

• Rewrite manifests

Not sure what your realtime requirement is, but having very frequent commits will unavoidably lead to a lot of snapshots that will need to be managed properly to avoid performance issues:

Having a high rate of commits produces data files, manifests, and snapshots which leads to additional maintenance. It is recommended to have a trigger interval of 1 minute at the minimum and increase the interval if needed.