Instead of creating geojson responses with HydroServer, try exposing spatial data to GeoServer. I see two possible ways to do this.
The first would require setting up a PostgreSQL container and connecting it to the Django application. Using a foreign table wrapper around the ODM2 SQLite files, it might be possible to create Postgres views based on the SQLite data. Using PostGIS, you could map spatial data from ODM2 to a PostGIS view. From there, it would be possible to create a PostGIS data store on GeoServer and publish a layer. This type of mapping could also be used to produce REFTs data without maintaining a separate catalog on HydroServer.
The second method would be to explore using SpatiaLite to map the data from ODM2 to GeoServer.
Both of these methods would allow data to be mapped directly from the SQLite files in iRODS to GeoServer web services without any data duplication. This would have to be tested fairly extensively as I don’t know what the performance implications would be for these services.
Instead of creating geojson responses with HydroServer, try exposing spatial data to GeoServer. I see two possible ways to do this.
The first would require setting up a PostgreSQL container and connecting it to the Django application. Using a foreign table wrapper around the ODM2 SQLite files, it might be possible to create Postgres views based on the SQLite data. Using PostGIS, you could map spatial data from ODM2 to a PostGIS view. From there, it would be possible to create a PostGIS data store on GeoServer and publish a layer. This type of mapping could also be used to produce REFTs data without maintaining a separate catalog on HydroServer.
The second method would be to explore using SpatiaLite to map the data from ODM2 to GeoServer.
Both of these methods would allow data to be mapped directly from the SQLite files in iRODS to GeoServer web services without any data duplication. This would have to be tested fairly extensively as I don’t know what the performance implications would be for these services.