[Radiance-general] sharing indirect values for parallel processing?

Hi Lars,

I believe that openmosix's omfs filesystems supports locking correctly across the networked nodes (open and fcntl). Also you should determine if you need to have dfsa enabled on top of omfs. It might be worth checking the documentation carefully. Also it looks like they are changing to another filesystem mechanism based on openGFS, although it looks like this is still in the works. However what does not work (or did not as of a year or so ago) are named pipes. These are needed if you are going to run multiple processes on each node and require sharing of memory (-PP). I think there are some tricks to work around this, but these essentially require starting and locking jobs on specific nodes.

You might want to consider a question to the openmosix listserv regarding file locking and named pipes. On the other hand you may just need to set up some tests and run the renderings to see what works and what does not.....


Lars O. Grobe wrote:



so after all I decided to use openmosix, start all processes on one machine and not to care about network file systems, as these processes read their data from the local fs before migrating to other clusters. This should work like using a SMP-machine.

But, to get it clear, I have to ask this once again (I read the chapter in RwR again): if I start processes on ONE machine (e.g. a SMP environment or an openmosix cluster) and all file access is on the local fs, will the sharing of ambient data work? The byte range locking would have to be done over the local file system, right? Or should I just use the -N option of rad (or the -PP of rpict?)?

TIA+CU Lars.

Radiance-dev mailing list