-
Notifications
You must be signed in to change notification settings - Fork 3.4k
Description
Hello,
I have an on premise Kubernetes cluster and I deployed CVAT according to this documentation: https://docs.cvat.ai/docs/administration/advanced/k8s_deployment_with_helm/
There are 2 points in the documentation that would be nice to improve.
Issue 1
In the last sections where you describe "How to use existing PersistentVolume...", I quickly realized that this PVC needs to have access mode "RWX", as it is going to be used by the following Pods:
- cvat-backend-server
- cvat-backend-worker-export
- cvat-backend-worker-import
- cvat-backend-worker-annotation
- cvat-backend-worker-utils
- cvat-backend-worker-chunks
It would be nice if you document this, as it is crucial for the creation of the PVC before attempting to deploy CVAT with Helm.
Issue 2
In the same section, the yaml of values you propose is erroneous. If you use the yaml you propose when CVAT is deployed an error message is shown that cvat-backend-worker-utils
and cvat-backend-worker-chunks
are not deployed. In order to fix that:
- The whole
utils
block should be indented by one tab more to the right. - A
chunks
block should be added.
This is what worked for me:
cvat:
backend:
permissionFix:
enabled: false
defaultStorage:
enabled: false
server:
additionalVolumes:
- name: cvat-backend-data
persistentVolumeClaim:
claimName: my-claim-name
worker:
export:
additionalVolumes:
- name: cvat-backend-data
persistentVolumeClaim:
claimName: my-claim-name
import:
additionalVolumes:
- name: cvat-backend-data
persistentVolumeClaim:
claimName: my-claim-name
annotation:
additionalVolumes:
- name: cvat-backend-data
persistentVolumeClaim:
claimName: my-claim-name
utils:
additionalVolumes:
- name: cvat-backend-data
persistentVolumeClaim:
claimName: my-claim-name
chunks:
additionalVolumes:
- name: cvat-backend-data
persistentVolumeClaim:
claimName: my-claim-name