Personal tools
You are here: Home Forum Extension of the platform Buffering calls to CORBA OMNIORB

Buffering calls to CORBA OMNIORB

Up to Extension of the platform

Buffering calls to CORBA OMNIORB

Posted by Samuel Woodhead at February 16. 2018

Hi All,

As part of a python exporter I've been trying to improve the performance of - I need to get all the face details of a mesh using the function GetElemFaceNodes of the SMESH Mesh Class. However, this has to be called for every volume element of potentialy huge meshes, and the call quickly becomes the bottleneck of the exporter speed.

I have tried to make parallel threaded calls to reduce function overhead, but this seems to have no impact on the time. I can't seem to find any documentation about CORBA multithreading calls either, or anything in the Salome documentation. I'm wondering if I need to open up multiple ports to the naming service to do this?

Any developers that have knowledge of how I can make efficiently buffered or parallel calls to the naming service? Any sort of settings in Salome that are required to make it possible?

Best Wishes,


Powered by Ploneboard
Document Actions