This framework is responsible for storing, managing and retrieving binary data. As it covers quite a range of functionality it is split into three layers which build on top of each other.
The lowest layer provides an API to store objects (chunks of binary data). The main entry point is ObjectStorage which provides access to the configured storage spaces. A space can be compared to buckets in the Amazon S3 terminology.
More details can be found in the Layer 1 documentation.
This layer uses the layer 1 to actually store the binary objects but also provides a way to store metadata in either a JDBC datasource or MongoDB.
The main entry point is BlobStorage which provides access to the individual BlobStorageSpace.
This layer is the central part of the storage framework as it provides utilities to reference blobs in entities, an HTTP dispatcher for efficient access and delivery as well as a conversion engine which provides variants of the stored blobs (e.g. a properly resized JPG image of a given EPS file).
More details can be found in the Layer 2 documentation.
The virtual file system (VFS) provides a unified interface for all parts of the system which either provide or consume files or file systems. Having a built-in FTP and SSH server, it supports external applications access via familiar protocols such as FTP, FTPS, SFTP, SCP. Besides, it can mount external file systems via FTP, FTPS, SFTP, SCP and CIFS to make external files available to internal sub systems.
Being a virtual file system, neither folders nor files have to exist anywhere physically. A VFSRoot can provide artificial folders and files and directly process the provided data.
A central root which is automatically provided is the L3Uplink of layer 2 which makes browsable spaces visible as folders in the VFS.
Note that the VFS also provides a management UI via http://localhost:9000/fs.
More details can be found in the Layer 3 documentation.
The util package provides some helper classes which are mainly used by the framework itself and in most cases shouldn't be accessed externally.
The s3 package provides a facility to manage and access Amazon S3 compatible object stores. In most cases it is advisable to either use the layer 1 or layer 2 as these provide additional functionalities such as replication, API emulation in the local file system or an efficient way of storing metadata.