Hackolade easily imports the schema from .parquet files, located on your local file system or on a shared directory, to represent the corresponding Entity Relationship Diagram and schema structure.  When multiple files are selected, you have the choice to either combine the schemas of the selected files (default), or to create a separate schema in the model for each selected file.



Parquet files can also be reverse-engineered from AWS S3, Azure Blob Storage, and Google Cloud Storage.



1) AWS S3

Give a meaningful name to the connection to identify it for later, and provide proper URI to your S3 bucket, and optional folder path.




If the S3 bucket is private, you must also provide authentication parameters (Access key id and Secret access key):


If you wish to handle AWS authentication through the credentials file, you may leave blank the Access Key ID and Secret Access Key fields, knowing that Hackolade supplies credentials following the recommendations described here.



2) Azure Blob Storage

Give a meaningful name to the connection to identify it for later, and provide proper Container name and Storage account name.



If the storage account is private, you must also provide your Storage access key:



If you wish to filter files, you may enter a file name prefix:


3) Google Cloud Storage

Give a meaningful name to the connection to identify it for later, and provide proper URI to your GCS bucket, and optional folder path.


If the  bucket is private, you must also access to the Private key: