Each data provider has a dedicated workspace with file system and database.
The workspace on file system is automatically created when a new account is set up.
The file hierarchy is predefined and must not be changed:
All instances on the platform automatically mount the data provider's workspaces, so a data provider's workspace is shared and accessible as a local directory by any application running on any instances (according to the permissions).
Applications access files stored in a data provider's workspace like files stored on local system.
A data provider can statically configure the location of the workspace in its application: /mnt/glusterfs/<PROVIDER_ID>/workspace
However, the location could change in future. It is recommended to check the location with the RESTful service provided by the ElasticFS API (authentication required)
A data provider can SSH to any instance of the platform using its credentials (the same than for accessing the portal)
Workspace is the home directory, so the data provider can maintain its applications and manage its data.
A data provider can use SCP to transfer files from/to its workspace directory. However, using SCP is more for installation and maintenance of the application.
To transfer files from/to the workspace, It is recommended to get a media link with the RESTful service provided by the Data Import API (authentication required). This service returns a media link for a protocol like FTP or FTPS and with a period of validity.
All data provider's workspaces are automatically backed-up with the file system by the ElasticFS each night.
Only administrators can restore the file system.
The database is not automatically created when a new account is set up.
A data provider must request by mail the creation of a dedicated database.
The following types of database are supported:
Please note that it is not possible to change the type of database once it is created. Data provider that want to migrate must manage and implement the migration process.
A technical user is automatically created with the database.
The default password for the dedicated technical user is given in the mail sent after the database has been created.
A data provider must use the dedicated technical user to manage its database.
The dedicated technical user has all privileges on the database's objects, so a data provider can manage its database, including schema, tables, etc.
A data provider must also use the dedicated technical user to manage other technical user privileges on its database's objects.
Data providers cannot create databases and users.
However, a data provider can create a read-only technical user with read-only privileges on its database with the RESTful service provided by the ElasticDB API (authentication required) A read-only user is useful to improve security in applications that must access the database in read-only.
The read-only technical user login and default password are the same.
It is recommended to change the default password of the technical user and read-only technical user as soon as possible (see maintenance)
All instances on the platform can access the database, so the data provider's database is shared and accessible by any application running on any instances.
By default, access is denied from the Internet. A data provider must request by mail authorization to access its dedicated database from the Internet.
Applications accesses the data provider's database with the PostgreSQL driver available in the application language.
A data provider can get connection information on its database with the RESTful service provided by the ElasticDB API. However, password is not returned. Data providers must know the password of its dedicated technical user (and the password of the optional dedicated read-only technical user)
A data provider can SSH to any instance of the platform using its credentials (the same than for accessing the portal), so he can use the psql tool to connect to its dedicated database and manage it.
Of course, data providers that have request access to their database from the Internet can use any tools like pgAdmin to manage their database.
A data provider can use any tool to export/import data from/to its database.
If done inside the platform (e.g. through a SSH session), data are exported/imported from files located on the data provider's workspace. The data provider can use SCP or Data Import API to transfer those files from/to the platform.
You can browse into your workspace by simple-clicking on the folders in the tree or double-clicking on folders in the file view.
The Workspace Exploration Tool is able to display certain kind of files such as images, text, and pdf files. However, you will need to use a compliant browser as Firefox or Chrome.