SFS file system

You can process SFS files by using the standard file operations such as read, write, rewrite, and delete.

Each SFS file has one internal primary index, which defines the physical ordering of the records in the file, and can have any number of secondary indexes, which provide alternate sequences in which the records can be accessed.

All data in SFS files is managed by an SFS server. SFS provides a system tool, sfsadmin, for performing administrative functions such as creating files and indexes, determining which volumes are available on the SFS server, and so on, through a command-line interface. For details, see the Encina publication in the related reference.

The SFS file system is nonhierarchical. That is, when identifying SFS files, you can specify only individual file names, not directory names, after the server name.

COBOL access to SFS files is nontransactional: each operation against an SFS file is atomic, that is, performed either in its entirety or not at all. In the event of an SFS system failure, the result of a file operation completed by a COBOL application might not be reflected in the SFS file.

The SFS file system conforms to 85 COBOL Standard.

With the SFS file system, you can easily read and write files to be shared with PL/I programs.