How Deadline Cloud works
With Deadline Cloud, you can create and manage rendering projects and jobs directly from digital content creation (DCC) pipelines and workstations.
You submit jobs to Deadline Cloud using the AWS SDK, AWS Command Line Interface (AWS CLI), or Deadline Cloud
job submitters. Deadline Cloud supports the Open Job Description (OpenJD)
for job template specification. For more information, see Open Job
Description
Deadline Cloud provides job submitters. A job submitter is a DCC plugin for submitting render jobs from a third-party DCC interface, such as Maya or Nuke. With a submitter, artists can submit rendering jobs from a third-party interface to Deadline Cloud where project resources are managed and jobs are monitored, all in one location.
With a Deadline Cloud farm, you can create queues and fleets, manage users, and manage project resource usage and costs. A farm consists of queues and fleets. A queue is where submitted jobs are located and scheduled to be rendered. A fleet is a group of worker nodes that run tasks to complete jobs. A queue must be associated with a fleet so that the jobs can render. A single fleet can support multiple queues and a queue can be supported by multiple fleets.
Jobs consist of steps, and each step consists of specific tasks. With the Deadline Cloud monitor, you can access statuses, logs, and other troubleshooting metrics for jobs, steps, and tasks.
Permissions in Deadline Cloud
Deadline Cloud supports the following:
-
Managing access to its API operations using AWS Identity and Access Management (IAM)
-
Managing access of workforce users using an integration with AWS IAM Identity Center
Before anyone can work on a project, they must have access to that project and the associated farm. Deadline Cloud is integrated with IAM Identity Center to manage workforce authentication and authorization. Users can be added directly to IAM Identity Center, or permission can be connected to your existing identity provider (IdP) such as Okta or Active Directory. IT administrators can grant access permissions to users and groups at different levels. Each subsequent level includes the permissions for the previous levels. The following list describes the four access levels from the lowest level to the highest level:
-
Viewer– Permission to see resources in the farms, queues, fleets, and jobs they have access to. A viewer can't submit or make changes to jobs.
-
Contributor– Same as a viewer, but with permission to submit jobs to a queue or farm.
-
Manager– Same as contributor, but with permission to edit jobs in queues they have access to, and grant permissions on resources that they have access to.
-
Owner– Same as manager, but can view and create budgets and see usage.
Note
These permissions don't give users access to the AWS Management Console or permission to modify Deadline Cloud infrastructure.
Users must have access to a farm before they can access the associated queues and fleets. User access is assigned to queues and fleets separately within a farm.
You can add users as individuals or as part of a group. Adding groups to a farm, fleet, or queue can make it easier to manage access permissions for large groups of people. For example, if you have a team that is working on a specific project, you can add each of the team members to a group. Then, you can grant access permissions to the entire group for the corresponding farm, fleet, or queue.
Software support with Deadline Cloud
Deadline Cloud works with any software application that can be run from a command line interface and controlled by using parameter values. Deadline Cloud supports the OpenJD specification for describing work as jobs with software script steps that are parameterized (such as across a frame range) into tasks. Assemble OpenJD job instructions into job bundles with Deadline Cloud tools and features to create, run, and license the steps from a third-party software application.
Jobs need licensing to render. Deadline Cloud offers usage-based-licensing (UBL) for a selection of software application licenses that is billed by the hour in minute increments based on usage. With Deadline Cloud, you can also use your own software licenses if you like. If a job can't access a license, it doesn't render and produces an error that displays in the task log in the Deadline Cloud monitor.
