HyperQueue support
HyperQueue is a tool designed to simplify execution of large workflows (task graphs) on HPC clusters. It allows you to execute a large number of tasks in a simple way, without having to manually submit jobs into batch schedulers. HEAppE Middleware directly supports HyperQueue as a task execution engine. This means that you can use HyperQueue to execute tasks on HPC clusters through HEAppE Middleware.
Warning
Prerequisity of using HyperQueue with HEAppE Middleware is to have Slurm workload manager
and HyperQueue
installed on HPC cluster as module.
To enable HyperQueue support in HEAppE Middleware, you need to set the SchedulerType
attribute in the configuration file to 8
(HQ with Slurm) in the Clusters section.
HEAppE Middleware uses the HyperQueue Auto Allocation feature to automatically allocate resources for task execution.
GPU nodes support
When using HyperQueue with HEAppE Middleware, you can also use GPU nodes for task execution. To enable GPU nodes support, you need to set the PlacementPolicy
attribute in the HEAppE Task Specification to gpus/nvidia
. Amount of the GPUs will be computed by the HEAppE Middleware based on the MaxCores
attribute in the HEAppE Task Specification.
CPU nodes support
CPU nodes are used by default, you do not need to set any additional attributes in the HEAppE Task Specification.
Note
When using HyperQueue, also please make sure that you are using the corrrect HPC partition / queue
for the task execution.