Command Server features

• cmdSrv executes faster than the old rsh based scp_spawn procedure.
• It is more secure than the old system (doesn't use r-services).
• Each server host machine has two control files:

  A .rhosts-like file controls which clients are 'trusted hosts'.

  A "commands" file controls which commands are accepted from clients.

• It is more reliable and easier to maintain and modify.
• Static load balancing is provided (you can have more than one server machine).

  You can refresh the server's list of supported client hosts without restarting the server.

  You can refresh the server's list of supported command classes without restarting the server.

  On the vms client side, you can change which unix machines handle which command classes by redefining logical names and don't need to restart the scp (see below for the meaning of the word 'class'.

• A flexible and portable API is provided for client programmers by calling cmdsrv_msg_sync(). The api makes it easy to send messages from standalones (example provided).

• An extensive log of transaction details is kept on disk for debugging purposes. The logging of transactions may be easily turned on or off by setting an environment variable.
• On VMS, the panel file button variables provide an easy to use and flexible interface to cmdSrv. The button variables indicate the unix command to execute (which may be a native command or script file of course), and may contain up to 7 command options.

Note:

This was a restriction on the 1.0 release that needs to be tested for 1.1... There is currently a restriction in the specification of unix command options. Only single word qualifiers starting with "-" are supported. (eg "-cfg filename" is NOT a supported command option). A workaround is to put the command with the two word option in a script file on unix and to execute that instead of executing the command directly.

Return to cmdSrv main page.