Skip to content

Google GKE


Google GKE is a managed Kubernetes cluster that allows the execution of containerized workloads in Google Cloud at scale.

Nextflow Tower offers native support for Google GKE clusters and streamlines the deployment of Nextflow pipelines in such environments.


You need to have a GKE cluster up and running.

Make sure you have followed the steps in the Cluster preparation guide to create the cluster resources required by Nextflow Tower.

Compute environment setup#

1. In a workspace choose "Compute environments" and then, click on the New Environment button.

2. Enter the Name for this environment, for example, My GKE.

3. Select Google GKE as the target platform.

4. Select your Google Cloud credentials. The credentials are needed to identify the user that will access the GKE cluster.

5. Select the Location where the GKE cluster is located.

Regional and zonal clusters

GKE clusters can be either regional or zonal. For example, us-west1 identifies the United States West-Coast region, which has three zones: us-west1-a, us-west1-b, and us-west1-c.
Tower self-completion only shows regions. You should manually edit this field if your GKE cluster was created by zone rather than regionally.

6. The field Cluster name lists all GKE clusters available in the selected location. Choose the one you want to use to deploy the Nextflow execution.

7. Specify the Kubernetes Namespace that should be used to deploy pipeline executions.

If you have followed the example in the cluster preparation guide, this field should be tower-nf.

8. Specify the Kubernetes Head service account that will be used to grant permissions to Tower to deploy the pods executions and related.

If you have followed the cluster preparation guide, this field should be tower-launcher-sa.

9. The Storage claim field allows you to specify the storage Nextflow should use as a scratch file system for the pipeline execution.

This should reference a Kubernetes persistence volume with ReadWriteMany capability. Check the cluster preparation guide for details.

Advanced options#

The following parameters are available:

1. The Storage mount path defines the file system path where the Storage claim is mount.

Default: /scratch

2. The Work directory field defines the file system path used as working directory by the Nextflow pipelines. It must be the same or a subdirectory of the Storage mount path at the previous point.

Default: the same as Storage mount path.

3. The Compute service account field allows you to specify the Kubernetes service account that the pipeline jobs should use. Default is the default service account in your Kubernetes cluster.


You now have everything you need to begin deploying massively scalable pipelines.

Jump to the documentation section for Launching Pipelines.