We currently utilize --tmp space in our srun/sbatch jobs and have a user trying to migrate to the slurmrestd interface for their work...

One issue they are finding is that they do not see a way to request --TMP from the API docs... although minimum_tmp_disk_per_node is listed... it doesn't appear to work on job submissions

Their attempt:
{
    "job": {

        "nodes": 1,
        "cpus_per_task": 1,
        "memory_per_node": 10,
        "time_limit": 1,
        "name": "test",
        "partition": "braintv",
        "batch_features": "minimum_tmp_disk_per_node=1GB"
    },
    "script": "#!/bin/bash\n sleep 100"
}

Is there a way to request TMP space in API jobs? If not... can I add it as a feature request?

Many thanks!