Scheduler
The component on the head node that decides which compute node runs each task, honoring the resources that task requested.
It places each task on a node that can satisfy the task’s request. A task asking for a GPU only runs where one exists, which is why a request nothing can satisfy waits indefinitely rather than failing.
Distributed compute
Related terms
- Head node: The machine in a cluster that coordinates work: it schedules tasks onto compute nodes and tracks their results.
- options(): The method that sets per-call resource requirements on a distributed function, such as `num_cpus` or `num_gpus`.
- Scheduling overhead: The fixed cost of dispatching a task and returning its result. When it approaches the cost of the task body, adding parallelism stops helping.
- Job scheduler: The queueing system on a traditional HPC cluster that decides when a submitted job runs. Eugo has no queue: you launch a cluster and dispatch work to it.
- Worker: A process on a compute node that executes tasks. Worker count, not node count, is what determines how many tasks run at once.