Tilebox Workflows

Tasks are the fundamental units of work within the Tilebox Workflow orchestrator. Each task represents a discrete operation or process that can be executed independently or as part of a larger workflow. Tasks are defined by their code, their inputs and their dependencies on other tasks. Implementing tasks involves defining the input parameters as well as specifying the action to be taken during its execution.

A job is a concrete execution of a workflow with specific input parameters. A job outlines the work to be achieved, comprising many tasks that may run in parallel or in sequence, depending on their dependencies. Submitting a job involves instantiating a root Task with concrete input parameters which then may trigger the execution of other tasks of the same workflow in the same job.

Task runners are the execution agents within the Tilebox Workflows ecosystem that carry out the tasks. Task runners can be deployed in different compute environments, including on-premise servers and cloud-based auto-scaling clusters. They are responsible for executing tasks as scheduled by the workflow orchestrator, ensuring that each task has the necessary resources and environment to run effectively.

Observability refers to a feature set within Tilebox Workflows that provides visibility into the execution of tasks and jobs. Observability tools, such as tracing and logging, allow users to track performance, diagnose issues, and gain insights into the operation of their jobs. This enables efficient troubleshooting and optimization of tasks and workflows.

Caches are shared data storage that can be used to store and retrieve data across tasks of a single job. Caches are used to store intermediate results, and to share data between tasks that are part of the same job. They are a key feature for enabling out-of-the-box distributed computing and reducing the need for redundant data processing.