Pivot tables in earth science
     
     * Data from satellites is divided into pieces (temporal, spatial, visual, etc.)
     * Pieces are stored in files.
     * Metadata about the files is stored in a database.
     * Files are available to minions in a cluster for processing.
     * Applications run on the minions, applying science algorithms,
          requesting and producing files.
     * New files are generated with new metadata, also made available
          for further processing.
     * Individual minions keep local caches of the files, and their metadata.
     * SQLite is used for maintaining the caches.
     * Portion of the sqlite schemas use the EAV model for file metadata.