-
Render Farm Blues
I’ve got Shake on all my systems, my Qmaster Admin has setup my Cluster, Compressor works fine but I seem to need to setup some kind of shared clustered storage.
I’m fibered to my SAN raids, I figured that if I set up my Shake scripts on the shared storage as well as having all the files etc existing on my SAN that I would be able to open the script from any workstation and work on the scripts, evidentially this needs to be done so that all workstations on the render farm have access to the same stuff to be able to be part of the farm. That doesn’t work, my FCP projects work fine that way but the archaic file structure of Shake can’t seem to figure out how to skip the local desktop in the directory pathway making opening scripts from the SAN on workstations other than the workstation that created the script unable to find the files.
Here’s what the manual says:
“When a rendering task is distributed using any render queue software, disk storage
must be shared by all the machines involved. This typically must be set up by the
system administrator via AFP, NFS, or other network facility.
Setting up and administering cluster storage is not trivial. Apple Qmaster offers a way
to mount the cluster storage device on all of the render nodes assigned to work on
the job, thereby making it easier to set up and confi gure the render farm.”Can someone translate this for me? I thought I was doing this by have a SAN system and having scripts and files all in the same place… I am unable to work on a script from other workstations on the SAN because it can’t find the files.
Thanks, Keith