Can Polyhedra operate without a file system?
Polyhedra is memory-based, and uses files to store snapshots of the database to provide data persistency. When Polyhedra needs to read from or write to files, it uses the standard Posix operating system calls to do so, and consequently does not depend on the specific hardware or filing system in use. For use on embedded systems where there may be no file system, or where the only persistent data storage (flash, battery-backed RAM, remote discs, etc.) requires special interface routines to be called, Polyhedra provides a variety of strategies: • it can be configured to read and write certain ‘files’ to a designated area of memory – which means that certain files can be ‘compiled and linked in’ to the Polyhedra DBMS; • it can be configured to read and write certain files via FTP, and so the files could be on a machine running FTP server software, to which the embedded machine has access via TCP/IP; and, • it can be configured to call a user-supplied routine when it wants to open a file;