Skip to content

README.md

Latest commit

 

History

History
200 lines (158 loc) · 8.44 KB

README.md

File metadata and controls

200 lines (158 loc) · 8.44 KB

Description

This module creates Kubernetes Persistent Volumes (PV) and Persistent Volume Claims (PVC) that can be used by a [gke-job-template].

gke-persistent-volume works with Filestore, Google Cloud Storage and Managed Lustre. Each gke-persistent-volume can only be used with a single file system so if multiple shared file systems are used then multiple gke-persistent-volume modules are needed in the blueprint.

NOTE: This is an experimental module and the functionality and documentation will likely be updated in the near future. This module has only been tested in limited capacity.

Example

The following example creates a Filestore and then uses the gke-persistent-volume module to use the Filestore as shared storage in a gke-job-template.

  - id: gke_cluster
    source: modules/scheduler/gke-cluster
    use: [network1]
    settings:
      master_authorized_networks:
      - display_name: deployment-machine
        cidr_block: <your-ip-address>/32

  - id: datafs
    source: modules/file-system/filestore
    use: [network1]
    settings:
      local_mount: /data

  - id: datafs-pv
    source: modules/file-system/gke-persistent-volume
    use: [datafs, gke_cluster]

  - id: job-template
    source: modules/compute/gke-job-template
    use: [datafs-pv, compute_pool, gke_cluster]

The following example creates a GCS bucket and then uses the gke-persistent-volume module to use the bucket as shared storage in a gke-job-template.

  - id: gke_cluster
    source: modules/scheduler/gke-cluster
    use: [network1]
    settings:
      master_authorized_networks:
      - display_name: deployment-machine
        cidr_block: <your-ip-address>/32

  - id: data-bucket
    source: modules/file-system/cloud-storage-bucket
    settings:
      local_mount: /data

  - id: datagcs-pv
    source: modules/file-system/gke-persistent-volume
    use: [data-bucket, gke_cluster]

  - id: job-template
    source: modules/compute/gke-job-template
    use: [datagcs-pv, compute_pool, gke_cluster]

The following example creates a Managed Lustre and then uses the gke-persistent-volume module to use the Lustre as shared storage in a gke-job-template.

  - id: gke_cluster
    source: modules/scheduler/gke-cluster
    use: [network1]
    settings:
      master_authorized_networks:
      - display_name: deployment-machine
        cidr_block: <your-ip-address>/32

  - id: data-managedlustre
    source: modules/file-system/managed-lustre
    settings:
      local_mount: /data

  - id: datalustre-pv
    source: modules/file-system/gke-persistent-volume
    use: [data-managedlustre, gke_cluster]

  - id: job-template
    source: modules/compute/gke-job-template
    use: [datalustre-pv, compute_pool, gke_cluster]

See example storage-gke.yaml blueprint for a complete example.

Authorized Network

Since the gke-persistent-volume module is making calls to the Kubernetes API to create Kubernetes entities, the machine performing the deployment must be authorized to connect to the Kubernetes API. You can add the master_authorized_networks settings block, as shown in the example above, with the IP address of the machine performing the deployment. This will ensure that the deploying machine can connect to the cluster.

Connecting Via Use

The diagram below shows the valid use relationships for the GKE Cluster Toolkit modules. For example the gke-persistent-volume module can use a gke-cluster module and a filestore module, as shown in the example above.

 graph TD;
 vpc--> |OneToMany| gke-cluster;
 gke-cluster--> |OneToMany| gke-node-pool;
 gke-node-pool--> |ManyToMany| gke-job-template;
 gke-cluster--> |OneToMany| gke-persistent-volume;
 gke-persistent-volume--> |ManyToMany| gke-job-template;
 vpc--> |OneToMany| filestore;
 vpc--> |OneToMany| gcs;
 vpc--> |OneToMany| managed-lustre;
 filestore--> |OneToOne| gke-persistent-volume;
 gcs--> |OneToOne| gke-persistent-volume;
 managed-lustre--> |OneToOne| gke-persistent-volume;
Loading

License

Copyright 2026 Google LLC

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

 http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Requirements

Name Version
terraform = 1.12.2
google >= 4.42
kubectl >= 1.7.0

Providers

Name Version
google >= 4.42
kubectl >= 1.7.0

Modules

No modules.

Resources

Name Type
kubectl_manifest.pv resource
kubectl_manifest.pvc resource
kubectl_manifest.pvc_namespace resource
google_client_config.default data source
google_container_cluster.gke_cluster data source

Inputs

Name Description Type Default Required
capacity_gib The storage capacity with which to create the persistent volume. number n/a yes
cluster_id An identifier for the GKE cluster in the format projects/{{project}}/locations/{{location}}/clusters/{{cluster}} string n/a yes
filestore_id An identifier for a filestore with the format projects/{{project}}/locations/{{location}}/instances/{{name}}. string null no
gcs_bucket_name The gcs bucket to be used with the persistent volume. string null no
labels GCE resource labels to be applied to resources. Key-value pairs. map(string) n/a yes
lustre_id An identifier for a lustre with the format projects/{{project}}/locations/{{location}}/instances/{{name}}. string null no
namespace Kubernetes namespace to deploy the storage PVC/PV string "default" no
network_storage Network attached storage mount to be configured. 
object({
    server_ip             = string,
    remote_mount          = string,
    local_mount           = string,
    fs_type               = string,
    mount_options         = string,
    client_install_runner = map(string)
    mount_runner          = map(string)
  })
 n/a yes
pv_name The name for PV. IF not set, a name will be generated based on the storage name. string null no
pvc_name The name for PVC. IF not set, a name will be generated based on the storage name. string null no

Outputs

Name Description
persistent_volume_claims An object describing the Kubernetes PersistentVolumeClaim created by this module.
pvc_name The name of the Kubernetes PVC created by this module.