1. Qube
2. Submitting a Cinema 4d job

Submitting a Cinema 4d job

Name
This is the name of the job of the job so it can be easily identified in the Qube! UI.

Priority
Every job in Qube is assigned a numeric priority. Priority 1 is higher than priority 100. This is similar to 1st place, 2nd place, 3rd place, etc. The default priority assigned to a job is 9999.

Instances
This is the number of copies of the application that will run at the same time across the network. The combination of "Instances=1" and "Max Instances=-1" means that this job will take as much of the farm as it can, and all jobs will share evenly across the farm.

Examples:

On a 12 slot(core) machine running Maya if you set
"Instances" to 4
"Reservations" to "host.processors=3"
Qube! will open 4 sessions of Maya on the Worker(s) simultaneously, which may consume all slots/cores on a given Worker.

if you set
"Instances" to 1
"Reservations" to "host.processors=1+"
Qube will open 1 session of Maya on a Worker, consuming all slots/cores ("host.processors=1+" is used for all slots/cores).

Max Instances
If resources are available, Qube! will spawn more than 'Instances' copies of the application, but no more than 'Max Instances'. The default of -1 means there is no maximum. If this is set to 0, then it won't spawn more than 'Instances' copies.

More on Instances & Reservations & SmartShare Studio Defaults

Range
Frame range for the job (e.g 1-100, or 1-100x3, or 1,3,7,10)

Most jobs require a frame range to execute on the workers. You can set this range in a few different ways :

• "1-100" will just render the range between 1 and 100
• "1-100x3" will render the range 1 to 100, every third frame, so 1, 4, 7, etc.
• "1,3,7,10" will only render the selected frames 1,3,7,10

Execution
How to break up frame range to be executed. Use QB_START_FRAME, QB_END_FRAME and QB_FRAME_NUMBER

When submitting a job to the farm it may be more efficient to "chunk" your job. This means that when the job is sent to the worker it tells the worker to render N consecutive frames before requesting more work. You would do this to keep from reopening the scene file for each frame. Large scene files can take substantial time to open, which is wasteful across dozens or hundreds of frames.

The drop down options are below:

• "Individual frames" this tells the worker to render 1 frame at a time.
• "Chunks with n frames" this tells the worker to render consecutively the number of frames specified in the field.
• "Split into n partitions" this tells the worker to render consecutively the total frames in the range divided by the number in the field.

Examples:

• range 1-100 with "individual frames" set will render 1 frame at a time
• range 1-100 with "Chunks with n frames" and the field set to 5 will send 20 frames to each instance
• range 1-100 with "Split into n partitions" and the field set to 4 will send 25 frames to each instance

rangeOrdering
Order to render the items. (Ascending=1,2,3,4,5...,Descending=10,9,8...,Binary=first,middle,last...)

You can set the order in which your frames are rendered. The drop down options are:

• "Ascending" - this will render the frames counting upwards from your start frame
• "Decending" - this will render the frames counting backwards from your end frame
• "Binary" - This will render the first, last, and middle frames of the range, then the middle frame of the first half and the middle frame of the second half, and so on. This is useful for sampling the frames in the sequence to make sure it is rendering correctly.

Use Preview Frames  

Enabling preview frames will create 2 jobs:

• A primary dependent job with a higher priority that will render the selected frames first 
• A secondary job with lower priority that will render the remaining frames. This will return the selected frames faster so that you can check the accuracy of your renders.

Frame Numbers

Choose the frames that you wish to render first. If left blank the default is to render the first frame, the last frame and the middle frame in that order. You can select specific frames by adding comma separated frame numbers e.g 1,2,10,15,75, or a range with, e.g., 1-100x5 (1 to 100, every 5th frame)

Preview Priority

Choose the priority for the preview job. This can be set by the site admin.

Preview Subjobs

Choose the number of instances / subjobs for the preview frames. By default, this is equal to the number of preview frames - that is, it will try to do all the preview frames at the same time.

Note that when you submit a job with preview frames enabled, it will actually submit 2 jobs—one with the preview frames list at a higher priority, and another with the rest of the agenda, at the normal priority (as specified in the job's Priority field). You will get, consequently, 2 job IDs for the submission.

New in 6.4-4
For applications/renderers that do not support using all cores while rendering (or changing that behavior, eg AfterEffects or 3dsMax), this section is not visible.

Render on all Cores
Checking this box means that once this job is assigned to a machine, no more jobs or instances will be assigned to the same machine until this job is complete. You might choose this if your know your render job is very memory intensive and shouldn't run alongside other jobs.

Min Free Slots
This is the number of slots that must be available on a worker in order for the worker to accept the job. For example, if you choose '1', a worker with 7 out of 8 cores already in use will still accept this job. However, if you were to choose '2', the same machine under the same circumstances would not accept the job. 

New in 6.4-4
For applications/renderers that do not support setting a specific number of threads, this section is not visible.

Slots = Threads
If this box is checked, it tells Qube and the application (eg, Maya) to use the specific number of threads listed in the "Specific Thread Count" field. If this is not checked, then the "Specific Thread Count" value is passed only to the application, but Qube is unaware of the number of cores/slots it should reserve. In most cases, you will want to check this box, unless you have Designer licenses, in which case you would only set the numeric value.

Specific Thread Count
This tells the renderer to use a specific number of threads for rendering. The default is one thread, which for any modern renderer will underperform, and when combined with the slots=threads, will swamp most workers by running as many instances as there are slots (eg, a 24 core machine would run 24 instances of the application/renderer). A better value is 8 (assuming you have 8 cores, as most modern machines do) which means that each instance of the job will use 8 threads to render, and, when combined with slots=threads, will reserve 8 slots while doing so.

Min File Size

Used to test the created output file to ensure that it is at least the minimum size specified. Put in the minimum size for output files, in bytes. A zero (0) disables the test.

regex_highlights

Used to add highlights into logs. Enter a regular expression that, if matched, will be highlighted in the information messages from stdout/stderr.

regex_errors

Used to catch errors that show up in stdout/stderr. For example, if you list "error: 2145" here and this string is present in the logs, the job will be marked as failed. This field comes pre-populated with expressions based on the application you are submitting from. You can add more to the list, one entry per line.

regex_outputPaths

Regular expression for identifying outputPaths of images from stdout/stder.  This is useful for returning information to the Qube GUI so that the "Browse Ouput" right-mouse menu works.

regex_progress

Regular expression for identifying in-frame/chunk progress from stdout/stderr. Used to identify strings for relaying the progress of frames.

regex_maxlines

Maximum number of lines to store for regex matched patterns for stdout/stderr. Used to truncate the size of log files.