Skip to main content

nextflow kuberun

Launch a Nextflow pipeline on a Kubernetes cluster.

Usage​

$ nextflow kuberun [options] [project]

Description​

The kuberun command extends the run command with a Kubernetes integration. It deploys the Nextflow runtime as a Kubernetes pod and requires the kubectl CLI. The kuberun command cannot run local Nextflow scripts. See Kubernetes for more information.

Options​

The kuberun command supports the following options from run:

  • -cache
  • -disable-jobs-cancellation
  • -dump-channels
  • -dump-hashes
  • -e.<key>=<value>
  • -entry
  • -h, -help
  • -hub
  • -latest
  • -main-script
  • -name
  • -offline
  • -params-file
  • -plugins
  • -preview
  • -profile
  • -qs, -queue-size
  • -resume
  • -r, -revision
  • -stub, -stub-run
  • -user
  • -with-conda
  • -with-dag
  • -N, -with-notification
  • -with-report
  • -with-spack
  • -with-timeline
  • -with-tower
  • -with-trace
  • -with-wave
  • -with-weblog
  • -w, -work-dir

The following new options are also available:

-head-cpus​

Number of CPUs requested for the Nextflow pod.

-head-image​

Container image for the Nextflow driver pod.

-head-memory​

Amount of memory requested for the Nextflow pod.

-head-prescript​

Script to run before the Nextflow pod starts.

-n, -namespace​

Kubernetes namespace to use.

-remoteConfig​

Add the specified file from the Kubernetes cluster to the configuration set.

-remoteProfile​

Configuration profile to use from -remoteConfig.

-v, -volume-mount​

Volume claim mounts. For example, my-pvc:/mnt/path.

Examples​

Run a pipeline on a Kubernetes cluster:

$ nextflow kuberun nextflow-io/hello
On this Page