Hi,

I'm planning to post a newly completed render queue
manager I wrote, which is for Houdini versions 8 and 9 running on
Linux to the Houdini Exchange in a few weeks, and wanted
to see if there is interest/curiousity or requests. It is called
ROXAN for “really simple but operational queuing
system for area networks (roqsan)”

Highlights:

• compatible and tested with H8 and H9 and mantra renderer in a 25 server facility
  written entirely in a small amount of Tcl/Tk script code (python was preferred but I had less experience with it and python was not integrated into H8)
  uses Firefox web browser for displays/control:
  server status, job status with image thumbnails,
  output logs
  uses a new ROXAN ROP DA for job submission
  does not require rsh or ssh - uses its own server process (which is launched as needed by xinetd daemon )
  instructions included for setup
  allows multiple CPUs per frame
  allows grouping of servers by show, resources, software, etc
  reasonably simple to install and use, or modify
  probably will be under the MIT license or similar
  
  
  I also plan to be available for support/help for anyone interested.
  
  -Dave Remba

Sounds cool - I'll keep an eye peeled!

Cheers,

J.C.

Great!

Looking forward to it!


kuba

daveremba
I'm planning to post a newly completed render queue
manager I wrote, which is for Houdini…

That's great Dave. Really looking forward to checking it out.

Great news Dave
More render queues for Hou8dini is great! More solutions for small studios.

Add my “Great News” to this topic as well!

I look forward to seeing what you come up with.

Really looking forward to checking it out.

me too, me too!

Hi,

Just a quick update on the new open source render queue manager
for Houdini on Linux - “ROXAN”

We've been testing it in production for a few weeks now, getting
the kinks out, and added a few new features by request

* ability to retry failed frames
* start mplay to play the rendered frames by clicking on the thumbnail
in the browser, or a “Play Frames” button in the DA
* pause/resume jobs
* schedule by round-robin or priority-sequential
* move jobs to a new render server group
* move jobs to a new prep (IFD/simulation) server
* checkpoint/restart the queue system for maintenance
* detect and reduce workload or restart overloaded servers

Timeline update for release/posting on the Exchange

probably two-three more weeks to finalize and finish
HTML documentation.

-Dave Remba