About mental queue

mental queue™ is a distributed process management system that can control all tasks of an entire production or computing service pipeline. It efficiently distributes large and varying numbers of possibly highly interdependent rendering and other tasks. Even the largest networks of machines and any number of users are supported. The unique architecture of mental queue enables node management across different operating systems and across interconnected production networks in different locations. Resources and tasks are dynamically allocated in the most efficient way possible at all times. In particular, mental queue permits optimal control of mental ray based rendering pipelines.

Job Queue Management

Jobs given to mental queue are placed in a prioritized queue. Job execution is handled by the two primary components of mental queue, namely the Manager and the Node. The Manager component is a daemon/service application which schedules job execution, splits each job into tasks and allocates tasks for processing. The Manager also provides web services to develop monitoring and auditing tools. An instance of the Node component runs on each computer assigned to the node pool and executes tasks given to it by the Manager. Job information is stored in an XML data structure, and job execution can be performed by external scripts, by dedicated software built within the scope of mental queue, or by plug-ins.

Manager Control

There are two mental queue components which can control the Manager, both of which can be installed on any computer on the network. These are the mental queue Monitor and Command Line Tool. The Monitor provides a graphical user interface to the manager and looks nearly identical on all platforms supported. The Command Line Tool provides non-graphical access and control. The user has the option of defining job priority, and can control whether fixed or dynamic task allocation is used. Fixed task allocation allows the user to define blocks of tasks to be processed together. If dynamic allocation is chosen, then the Manager automatically adjusts task groups according to Node performance.