Administering REXEC

Administering REXEC

Many commercially available transport protocols provide remote execution capabilities that are tied to the protocols. In UnixWare 7, the rlogin command, for example, allows remote users to log into a host on a TCP/IP network. REXEC is a remote execution facility that is independent of transport protocol. It allows a user to execute a process on a remote host, independent of the transport provider.

To the user, remote execution of a process through REXEC is transparent. Once the user executes a process on a remote machine, the user interfaces with the process as if it were running locally. However, all files referenced by a remote command are relative to the server. For example, if a user executes the who command remotely, a list of users logged in to the remote machine is displayed, not a list of users on the local machine.

For the system administrator, REXEC makes it possible to administer a set of machines remotely from a single location, without the overhead of logging in to the remote machines. REXEC gives administrators the ability to:

monitor the system activity of several machines from a single machine
stop and start network applications, such as the Network File System (NFS)
tune the performance of a set of machines by specifying that certain processes are to execute on a less active system

REXEC is service-based. A server defines services that a client can execute remotely. By default, when REXEC is first installed, a server has the following standard services defined:

rx: A service that allows a user on a client to execute a command or shell script on the server.
rl: A service that allows a user on a client to log in to the server.
rquery: A service that allows a client user to list the services available on a server for remote execution.

General instructions for invoking an REXEC service, as well as specific instructions for using the standard services, appear on the rexec(1bnu) manual page.