Browse Prior Art Database

Optimized Generic Remote Procedure Call Stubs for Distributed Computing Environment Core Services

IP.com Disclosure Number: IPCOM000118744D
Original Publication Date: 1997-Jun-01
Included in the Prior Art Database: 2005-Apr-01
Document File: 2 page(s) / 42K

Publishing Venue

IBM

Related People

Lin, D: AUTHOR [+3]

Abstract

In IBM AIX* Distributed Computing Environment (DCE) 1.3, the Remote Procedure Call (RPC) stubs for all interfaces between DCE daemons, control programs runtime in the Security, Directory, Time, and RPC services contain single call statement to the generic client/server stub routine in the runtime. The generic client and server stub routines are shared by all RPC calls. This significantly reduces the Random Access Memory/Direct Access Storage Device (RAM/DASD) requirements of DCE since multiple complex stubs compiled and linked to the daemons, control programs, and libraries are now very small.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 70% of the total text.

Optimized Generic Remote Procedure Call Stubs for Distributed Computing
Environment Core Services

      In IBM AIX* Distributed Computing Environment (DCE) 1.3, the
Remote Procedure Call (RPC) stubs for all interfaces between DCE
daemons, control programs runtime in the Security, Directory, Time,
and RPC services contain single call statement to the generic
client/server stub routine in the runtime.  The generic client and
server stub routines  are shared by all RPC calls.  This
significantly reduces the Random Access Memory/Direct Access Storage
Device (RAM/DASD) requirements of DCE since multiple complex stubs
compiled and linked to the daemons, control programs, and libraries
are now very small.

      The generic stubs can be shared not only by DCE components but
all DCE RPC applications.  In order to make them work for any
interfaces with any data types, the generic stubs are made a union of
all possible  stubs with all possible functions.  A number of control
parameters are  used to specialize the generic stubs at runtime.  As
such, the generic  RPC stub routines are not as efficient as the
ordinary stubs. Although  the overhead is only a small portion of the
entire RPC execution time,  it is worthwhile to improve the
performance of the heavy activities occurring between DCE core
components in a DCE cell.

      An important idea of this performance optimization with the
RAM/DASD consideration is to reverse the stub generalization process,
a ste...