Get started with the Batch Rendering service
The Azure Batch Rendering service offers cloud-scale rendering capabilities on a pay-per-use basis. The Batch Rendering service handles job scheduling and queueing, managing failures and retries, and auto-scaling for your render jobs. The Batch Rendering service supports rendering apps including Autodesk Maya, 3ds Max, Arnold, and V-Ray. The Batch plug-in for Maya 2017 makes it easy to start a rendering job on Azure right from your desktop.
With Maya and 3ds Max, you can run jobs using the BatchLabs desktop application or the Batch Templates CLI. Using the Azure Batch CLI, you can run Batch jobs without writing code. Instead, you can use template files to create Batch pools, jobs, and tasks. For more information, see Use Azure Batch CLI Templates and File Transfer.
The Batch Rendering service currently supports the following applications:
On CentOS 7 rendering nodes:
- Autodesk Maya I/O 2017 Update 5 (cut 201708032230)
- Autodesk Maya I/O 2018 Update 2 cut 201711281015
- Autodesk Arnold for Maya 2017 (Arnold version 220.127.116.11) MtoA-18.104.22.168-2017
- Autodesk Arnold for Maya 2018 (Arnold version 22.214.171.124) MtoA-126.96.36.199-2018
- Chaos Group V-Ray for Maya 2017 (version 3.60.04)
- Chaos Group V-Ray for Maya 2018 (version 3.60.04)
- Blender (2.68)
On Windows Server 2016 rendering nodes:
- Autodesk Maya I/O 2017 Update 5 (version 17.4.5459)
- Autodesk Maya I/O 2018 Update 2 (version 188.8.131.5276)
- Autodesk 3ds Max I/O 2018 Update 4 (version 184.108.40.20654)
- Autodesk Arnold for Maya (Arnold version 220.127.116.11) MtoA-18.104.22.168-2017
- Autodesk Arnold for Maya (Arnold version 22.214.171.124) MtoA-126.96.36.199-2018
- Autodesk Arnold for 3ds Max (Arnold version 188.8.131.52 )(version 1.2.926)
- Chaos Group V-Ray for Maya (version 3.52.03)
- Chaos Group V-Ray for 3ds Max (version 3.60.02)
- Blender (2.79)
To use the Batch Rendering service, you need:
- Azure account.
- Azure Batch account. For guidance on creating a Batch account in the Azure portal, see Create a Batch account with the Azure portal.
- Azure Storage account. The assets used for your rendering job are typically stored in Azure Storage. You can create a storage account automatically when you set up your Batch account. You can also use an existing storage account. For storage account options in Batch, see the Batch feature overview.
- Environment variables. If your solution modifies environment variables, make sure that the values of
AZ_BATCH_SOFTWARE_ENTITLEMENT_TOKENare kept intact and present when any of the licensed applications above is called. Otherwise, you are likely to encounter software activation issues.
- BatchLabs (optional). BatchLabs is a free, rich-featured, standalone client tool to help create, debug, and monitor Azure Batch applications. While not required to use the Rendering service, it is a useful option to develop and debug your Batch solutions.
To use the Batch plug-in for Maya, you need:
- Autodesk Maya 2017.
- A supported renderer such as Arnold for Maya or V-Ray for Maya.
Basic Batch concepts
Before you begin using the Batch Rendering service, it's helpful to be familiar with a few Batch concepts, including compute nodes, pools, and jobs. To learn more about Azure Batch in general, see Run intrinsically parallel workloads with Batch.
Batch is a platform service for running compute-intensive work, like rendering, on a pool of compute nodes. Each compute node in a pool is an Azure virtual machine (VM) running either Linux or Windows.
For more information about Batch pools and compute nodes, see the Pool and Compute node sections in Develop large-scale parallel compute solutions with Batch.
A Batch job is a collection of tasks that run on the compute nodes in a pool. When you submit a rendering job, Batch divides the job into tasks and distributes the tasks to run on the compute nodes in the pool.
You can use the Azure Portal to monitor jobs and diagnose failed tasks by downloading application logs and by remotely connecting to individual VMs using RDP or SSH. You can also manage, monitor, and debug using the BatchLabs tool.
For more information about Batch jobs, see the Job section in Develop large-scale parallel compute solutions with Batch.
Options for provisioning required applications
Multiple applications might be required to render a job, for example, the combination of Maya and Arnold, or 3ds Max and V-Ray, as well as other third-party plugins, if required. Additionally, some customers might require specific versions of these applications. As such, there are several methods available for provisioning the required applications and software:
Pre-configured VM images
Azure provides Windows and Linux images each with a single version of Maya, 3ds Max, Arnold, and V-Ray pre-installed and ready to use. You can select these images in the Azure portal, the Maya plug-in, or BatchLabs when you create a pool.
In the Azure portal and in BatchLabs, you can install one of the VM images with the pre-installed applications as follows: In the Pools section of your Batch account, select New, then in Add Pool, select Graphics and Rendering (Linux/Windows) from the Image type drop-down list:
Scroll down and under Graphics and rendering licensing click Select software and pricing. Choose one or more of the software licenses:
The specific license versions provided match the versions in the "Supported Applications" section above.
Azure Batch allows you to provide your own custom image. Using this option, you can configure your VM with the exact applications and specific versions that you require. For more information see Use a custom image to create a pool of virtual machines. Note that Autodesk and Chaos Group have modified Arnold and V-Ray respectively to validate against our own licensing service. You will need to ensure you have the versions of these applications with this support, otherwise the pay-per-use licensing won't work. This license validation isn't required for Maya or 3ds Max as the current published versions don't require a license server when running headless (in batch/command-line mode). Please contact Azure support if you're not sure how to proceed with this option.
Options for submitting a render job
Depending on the 3D application you use, there are various options for submitting render jobs to the service:
With Maya, you can use:
With 3ds Max, you can use:
- BatchLabs desktop application (see BatchLabs-data for guidance on using 3ds Max BatchLabs templates)
- Batch Templates CLI
The 3ds Max Batch Labs templates allow you to render VRay and Arnold scenes using the Azure Batch Rendering Service. There are two variations of the template for VRay and Arnold, one for standard scenes and one for more complex scenes that require a 3ds Max path file to assets and textures (.mxp file). For more information on the 3ds Max Batch Labs templates, see the BatchLabs-data repo on GitHub.
Additionally, you can use the Batch Python SDK to integrate the rendering service with your existing pipeline.
Use the Batch plug-in for Maya to submit a render job
With the Batch plug-in for Maya, you can submit a job to the Batch Rendering service right from Maya. The following sections describe how to configure the job from the plug-in and then submit it.
Load the Batch plug-in for Maya
The Batch plug-in is available on GitHub. Unzip the archive to a directory of your choice. You can load the plug-in directly from the azure_batch_maya directory.
To load the plug-in in Maya:
- Run Maya.
- Open Window > Settings/Preferences > Plug-in Manager.
- Click Browse.
- Navigate to and select azure_batch_maya/plug-in/AzureBatch.py.
Authenticate access to your Batch and Storage accounts
To use the plug-in, you need to authenticate using your Azure Batch and Azure Storage account keys. To retrieve your account keys:
- Display the plug-in in Maya, and select the Config tab.
- Navigate to the Azure portal.
- Select Batch Accounts from the left-hand menu. If necessary, click More Services and filter on Batch.
- Locate the desired Batch account in the list.
- Select the Keys menu item to display your account name, account URL, and access keys:
- Paste the Batch account URL into the Service field in the Batch plug-in.
- Paste the account name into the Batch Account field.
- Paste the primary account key into the Batch Key field.
- Select Storage Accounts from the left-hand menu. If necessary, click More Services and filter on Storage.
- Locate the desired Storage account in the list.
- Select the Access Keys menu item to display the storage account name and keys.
- Paste the Storage account name into the Storage Account field in the Batch plug-in.
- Paste the primary account key into the Storage Key field.
- Click Authenticate to ensure that the plug-in can access both accounts.
Once you have successfully authenticated, the plug-in sets the status field to Authenticated:
Configure a pool for a render job
After you have authenticated your Batch and Storage accounts, set up a pool for your rendering job. The plug-in saves your selections between sessions. Once you've set up your preferred configuration, you won't need to modify it unless it changes.
The following sections walk you through the available options, available on the Submit tab:
Specify a new or existing pool
To specify a pool on which to run the render job, select the Submit tab. This tab offers options for creating a pool or selecting an existing pool:
- You can auto provision a pool for this job (the default option). When you choose this option, Batch creates the pool exclusively for the current job, and automatically deletes the pool when the render job is complete. This option is best when you have a single render job to complete.
- You can reuse an existing persistent pool. If you have an existing pool that is idle, you can specify that pool for running the render job by selecting it from the dropdown. Reusing an existing persistent pool saves the time required to provision the pool.
- You can create a new persistent pool. Choosing this option creates a new pool for running the job. It does not delete the pool when the job is complete, so that you can reuse it for future jobs. Select this option when you have a continuous need to run render jobs. On subsequent jobs, you can select reuse an existing persistent pool to use the persistent pool that you created for the first job.
Specify the OS image to provision
You can specify the type of OS image to use to provision compute nodes in the pool on the Env (Environment) tab. Batch currently supports the following image options for rendering jobs:
Choose a VM size
Specify licensing options
You can specify the licenses you wish to use on the Env tab. Options include:
- Maya, which is enabled by default.
Arnold, which is enabled if Arnold is detected as the active render engine in Maya.
If you wish to render using your own license, you can configure your license end point by adding the appropriate environment variables to the table. Be sure to deselect the default licensing options if you do so.
You are billed for use of the licenses while VMs are running in the pool, even if the VMs are not currently being used for rendering. To avoid extra charges, navigate to the Pools tab and resize the pool to 0 nodes until you are ready to run another render job.
Manage persistent pools
You can manage an existing persistent pool on the Pools tab. Selecting a pool from the list displays the current state of the pool.
From the Pools tab, you can also delete the pool and resize the number of VMs in the pool. You can resize a pool to 0 nodes to avoid incurring costs in between workloads.
Configure a render job for submission
Once you have specified the parameters for the pool that will run the render job, configure the job itself.
Specify scene parameters
The Batch plug-in detects which rendering engine you're currently using in Maya, and displays the appropriate render settings on the Submit tab. These settings include the start frame, end frame, output prefix, and frame step. You can override the scene file render settings by specifying different settings in the plug-in. Changes you make to the plug-in settings are not persisted back to the scene file render settings, so you can make changes on a job-by-job basis without needing to reupload the scene file.
The plug-in warns you if the render engine that you selected in Maya is not supported.
If you load a new scene while the plug-in is open, click the Refresh button to make sure the settings are updated.
Resolve asset paths
When you load the plug-in, it scans the scene file for any external file references. These references are displayed in the Assets tab. If a referenced path cannot be resolved, the plug-in attempts to locate the file in a few default locations, including:
- The location of the scene file
- The current project's sourceimages directory
- The current working directory.
If the asset still cannot be located, it is listed with a warning icon:
If you know the location of an unresolved file reference, you can click the warning icon to be prompted to add a search path. The plug-in then uses this search path to attempt to resolve any missing assets. You can add any number of additional search paths.
When a reference is resolved, it is listed with a green light icon:
If your scene requires other files that the plug-in has not detected, you can add additional files or directories. Use the Add Files and Add Directory buttons. If you load a new scene while the plug-in is open, be sure to click Refresh to update the scene's references.
Upload assets to an asset project
When you submit a render job, the referenced files displayed in the Assets tab are automatically uploaded to Azure Storage as an asset project. You can also upload the asset files independently of a render job, using the Upload button on the Assets tab. The asset project name is specified in the Project field and is named after the current Maya project by default. When asset files are uploaded, the local file structure is preserved.
Once uploaded, assets can be referenced by any number of render jobs. All uploaded assets are available to any job that references the asset project, whether or not they are included in the scene. To change the asset project referenced by your next job, change the name in the Project field in the Assets tab. If there are referenced files that you wish to exclude from uploading, unselect them using the green button beside the listing.
Submit and monitor the render job
After you have configured the render job in the plug-in, click the Submit Job button on the Submit tab to submit the job to Batch:
You can monitor a job that is in progress from the Jobs tab on the plug-in. Select a job from the list to display the current state of the job. You can also use this tab to cancel and delete jobs, as well as to download the outputs and rendering logs.
To download outputs, modify the Outputs field to set the desired destination directory. Click the gear icon to start a background process that watches the job and downloads outputs as it progresses:
You can close Maya without disrupting the download process.
To learn more about Batch, see Run intrinsically parallel workloads with Batch.