azure-batch-cli-extensions/samples/blender

Azure Batch Blender

This sample shows how to use Blender to render using a parameterized template that specifies a virtualMachineConfiguration based autopool.

Features used by this sample

• Job template with parameterization
• Parametric sweep task factory
• Automatic persistence of task output files to Azure Storage

Prerequisites

You will need an Azure Batch account with a linked Azure Storage account. See Create an Azure Batch account using the Azure portal for details.

You will need a Blender scene to render.

Upload files

To upload your blender files:

az batch file upload --local-path <path> --file-group blender-data

Run this command on a folder containing the blender files you want to process. blender-data is the default value of the sceneData parameter in the job template. If you upload your files to a different file group, be sure to provide this value for the sceneData parameter when creating your job (see the next section).

Create the job

Modify the parameters.json file to specify appropriate parameters for the job. The full set of available parameters is defined in render.json:

Parameter	Required	Description
blendFile	Mandatory	File name of the Blender scene to be rendered
sceneData	Optional	Name of the file group where the input data is stored. Defaults to blender-data.
numberNodes	Optional	Number of nodes in the Azure Batch pool where the job will run. Defaults to 5.
vmSize	Optional	Size of the virtual machines that run the application. Defaults to STANDARD_A1.
jobName	Mandatory	Prefix of the name of the Azure Batch job, also used to prefix rendered outputs.
frameStart	Mandatory	Index of the first frame to render.
frameEnd	Mandatory	Index of the last frame to render.
outputFileStorageUrl	Mandatory	SAS URL for a container where outputs will be stored.

When you are ready to run the job, use this command:

az batch job create --template render.json --parameters parameters.json

This job uses an autopool which will automatically be deleted once the job reaches the Completed state. If the job never reaches the Completed state (e.g. when tasks are unable to run because of scheduling errors, or errors downloading files from storage), you will continue to be charged for the pool. In this case, you may want to use the Azure portal to manually delete the pool to ensure you're not billed unnecessarily.

Monitor the job

You can use this command to monitor the tasks in the job and their progress:

az batch task list --job-id <jobid>`

You can also use the Azure portal or Batch Explorer for monitoring.

Structure of the sample

File	Content
render.json	Specifies the job to run, including a parametric sweep task factory, the autopool definition, parameter and variable definitions.
parameters.json	Provides parameter values used to create the actual job that will run.