CLI reference

ewoksdask local

Start a Dask cluster with local machine resources.

usage: ewoksdask local [-h] [--n-workers N_WORKERS] [--no-processes]
                       [--threads-per-worker THREADS_PER_WORKER]
                       [--scheduler-port SCHEDULER_PORT]
                       [--dashboard-address DASHBOARD_ADDRESS]
                       [--worker-dashboard-address WORKER_DASHBOARD_ADDRESS]
                       [-o OPTION=VALUE]
                       [-l {debug,info,warning,error,critical}]

Named Arguments

--n-workers

Default: 1

--no-processes

Disable multiprocessing

Default: True

--threads-per-worker

--scheduler-port

--dashboard-address

Default: “:8787”

--worker-dashboard-address

-o, --option

Other options: https://distributed.dask.org/en/latest/api.html#distributed.LocalCluster

Default: []

-l, --log

Possible choices: debug, info, warning, error, critical

Log level

Default: info

ewoksdask slurm

Start a Dask cluster with Slurm resources.

usage: ewoksdask slurm [-h] [--n-workers N_WORKERS]
                       [--minimum-jobs MINIMUM_JOBS]
                       [--maximum-jobs MAXIMUM_JOBS] [--queue QUEUE]
                       [--cores CORES] [--gpus GPUS] [--processes PROCESSES]
                       [--memory MEMORY] [--walltime WALLTIME]
                       [-o OPTION=VALUE]
                       [-l {debug,info,warning,error,critical}]

Named Arguments

--n-workers

--minimum-jobs

Default: 0

--maximum-jobs

Default: 0

--queue

--cores

Default: 2

--gpus

--processes

Default: 1

--memory

Default: “1GB”

--walltime

Default: “01:00:00”

-o, --option

Other options: https://jobqueue.dask.org/en/latest/generated/dask_jobqueue.SLURMCluster.html

Default: []

-l, --log

Possible choices: debug, info, warning, error, critical

Log level

Default: info