Overview of Pipeline Phases
Init (Initialisation) Phase
The pipeline takes a CSV (or similar) input file and creates the necessary directories and config files for the pipeline to being running.
Scan Phase
Second phase of the pipeline involves scanning a subset of the NetCDF/HDF/Tiff files to determine certain parameters:
Ensure NetCDF/HDF/Tiff files can be converted successfully using one of the available drivers:
Calculate expected memory (for job allocation later.)
Calculate estimated chunk sizes and other values.
Determine file-type (JSON or Parquet) for final Kerchunk file.
Identify Identical/Concat dims for use in Compute phase.
Determine any other specific parameters for the dataset on creation and concatenation.
Compute Phase
Building the Kerchunk file for a dataset requires a multi*step process:
Create Kerchunk references for each archive-type file.
Save cache of references for each file prior to concatenation.
Perform concatenation (abort if concatenation fails, can load cache on second attempt).
Perform metadata corrections (based on updates and removals specified at the start)
Add Kerchunk history global attributes (creation time, pipeline version etc.)
Reconfigure each chunk for remote access (replace local path with https:// download path)
Validation Phase
Kerchunk files must be validated against equivalent Xarray objects from the original NetCDF:
Ensure all variables present in original files are present in Kerchunk (barring exceptions)
Ensure array shapes are consistent across Kerchunk/NetCDF
Ensure data representations are consistent (values in array subsets)
Several options and switches can be configured for the validation step, see the BypassSwitch class.
Next Steps
Kerchunk files that have been validated are moved to a complete
directory with the project code as the name, plus the kerchunk revision krX.X. These can then be linked to a catalog or ingested into the CEDA archive where appropriate.