[ace-users] CSD with reliable oneway calls

[Douglas C. Schmidt]

Hi Jochen,

   Thanks very much for using the PRF.  The CSD was provided by OCI,
so someone there should be able to help!

Doug

>Subject: TAO : CSD with reliable oneway calls
>
>    ACE VERSION: 5.5.6
>
>    HOST MACHINE and OPERATING SYSTEM:
>      sun solaris 10
>
>    TARGET MACHINE and OPERATING SYSTEM, if different from HOST:
>    COMPILER NAME AND VERSION (AND PATCHLEVEL):
>            
>      CC: Sun C++ 5.7 Patch 117830-08 2006/07/12
>
>    THE $ACE_ROOT/ace/config.h FILE [if you use a link to a platform-
>    specific file, simply state which one]:
>
>      #include "ace/config-sunos5.10.h"
>    
>    THE $ACE_ROOT/include/makeinclude/platform_macros.GNU FILE [if you
>    use a link to a platform-specific file, simply state which one
>    (unless this isn't used in this case, e.g., with Microsoft Visual
>    C++)]: 
>    
>      include $(ACE_ROOT)/include/makeinclude/platform_sunos5_sunc++.GNU
>
>    CONTENTS OF $ACE_ROOT/bin/MakeProjectCreator/config/default.features
>    (used by MPC when you generate your own makefiles):
>      ssl=0
>      qos=0
>
>
>    DOES THE PROBLEM AFFECT:
>        EXECUTION:
>        MultiThreading - CSD - Oneway calls
>
>    SYNOPSIS:
>       Possible bypass of CSD Strategy setting SyncScope Policy ?      
>
>
>    DESCRIPTION:
>     During performance tests we observed that there were oneway calls 
>missed. So we decided to introduce 'reliable oneway calls using the 
>SyncScope Policy'.
>     We are using the CSD Framework by applying our own strategy to the 
>child poa. This strategy puts all calls into a queue for later 
>dispatching in our process framework. This main application is single 
>threaded and dispatches events of several input queues sequentially.
>     Because this application is not thead-safe it is important for us 
>that all CORBA calls are dispatched using our strategy.
>     But after setting the reliable oneway calls at ORB-level we observed 
>that some calls are are not executed in the main thread. Removing the 
>policy setting for oneway calls avoided this problem. 
>     Is there a possibility that the strategy changes it's behaviour 
>comparable to the behaviour of a 'regular' orb?
>
>
>     Best regards,
>       Jochen Rothenbacher


-- 
Dr. Douglas C. Schmidt                       Professor and Associate Chair
Electrical Engineering and Computer Science  TEL: (615) 343-8197
Vanderbilt University                        WEB: www.dre.vanderbilt.edu/~schmidt
Nashville, TN 37203                          NET: d.schmidt at vanderbilt.edu