What the batch is this about?
Batch processing is the running of jobs without user interaction, where the processing happens on a bounded data set, that have already been stored over a period of time.
A job is the way of referring to a particular execution, in this context.
Batch jobs are typically long-term, heavyweight processes requiring on-demand provisioning of computational resources.
BatchX at a glance
BatchX is a fully managed service for easily running batch jobs in the cloud.
At a glance, it allows you to:
- Upload files to a scalable file-system.
- Import the job definitions as Docker images, aimed at processing input files.
- Run the jobs at scale, with the desired amount of vCPUs, RAM and GPUs.
- Download the results stored in the scalable file system or use them as inputs for new executions.
- Easily manage all the asset generated (files, images, jobs).
BatchX run-time model offers a complete decoupling between the image implementation and the platform infrastructure (see The BatchX contract), allowing you to just focus on solving your business needs, without worrying about infrastructure concerns like scaling, patching, security and reliability.