...
Downloading and importing grid data can be time consuming, and the data can take up much space in the database. If one is connected to the archive (and will be the next time the data is to be used), it is also possible to only import a reference to the data by selecting the tick box 'Import import grids as reference.' This requires the user, however, to maintain access to that file for the data to be used later. This option is available for the 'search and download datasets' and 'search and download events' panel. Once the user decides to import the grid data, the data will not be downloaded but only a reference to where the data can be found in the archive will be is stored. When the data is needed in the FEWS system, for example when the user wants to visualize the data or use it in a workflow, this data is downloaded (streamed) by using the stored reference. The data will be stored in the local cache and is not be stored into the databasein the database.
N.B. The actual download and import process works as follows. Upon pressing the 'download data' or 'download and import data' button, FEWS will download the relevant metadata.xml and runinfo.xml files from the archive, and place a dummy (0 byte) netCDF file next to it. In the import process, FEWS will recognize that the data is a dummy, and then tries to import the reference by asking the archive for the correct path to the data.
Importing states
The states can be downloaded and imported as a part of ‘simulated forecast’ data set, together with simulated time series, reports and modifiers. For this purpose an item ‘simulated forecast’ should be selected from the data set drop down list.
...
6) After step 5 the button "Upload to archive" will enabled. Note that when you change the selected time series the button will be unavailable again and that you should check the availability of the data again. When you press the button "Upload to archive" the selected time series will be uploaded to archive. For the periods for which the time series are not already archived to time series will be exported to netcdf and uploaded to the archive. For the periods for which the data is already archived only the edited values will be uploaded in pi-xml format to the archive. Both uploads are handled by the webservices of the archive. To insert the edited data into the archive a workflow in FEWS must be executed. To insert the newly added netcdf-files into the archive a task at the archive server must be executed. The upload process will determine automatically to which areas the time series belong. This area is determined by the value of a configured location attribute id. It is possible to configure more than one location attribute id. If a location belongs to multiple areas the new data will be added to all areas.
...