Help with storageclass / persistence volumes on BYOK

Hi everyone !

Here is some context:

  • I’ve deployed a custom Kubernetes cluster (1.29.2) on a scaleway ElasticMetal (not VPS instances), to be used as BYOK with qovery

  • on that cluster I then don’t have access to scaleway csi volumes: Managing Block Storage volumes with the Scaleway CSI as it don’t match scaleway csi pre-requisites:

  • Created a Kubernetes cluster running on Scaleway Instances (v1.21+).

→ Not my case, I use ElascticMetal machine != Scaleway instances.

  • I have currently no pv / storageclass on my cluster:
kubectl get pv
No resources found

kubectl get storageclass
No resources found
  • I then deploy a qovery database on that cluster (using postgresql), it logically run into a persistence volume error:
🛰 Database at new version 15 has 1 pods: 0 running, 1 starting, 0 terminating and 0 in error
┃  |__ Pod postgres-zbd790d29-test-0 is STARTING
┃     |__ ⚠️ 0/1 nodes are available: pod has unbound immediate PersistentVolumeClaims. preemption: 0/1 nodes are available: 1 Preemption is not helpful for scheduling.
┃
┃ 💽 Network volume data-postgres-zbd790d29-test-0 is STARTING
┃  |__ ⚠️ storageclass.storage.k8s.io "scw-sbv-ssd-0" not found

→ My goal is to have qovery database to be able to allocate persistent volume on my BYOK cluster:

:one: Does the storageclass required by qovery database is fixed to “scw-sbv-ssd-0” ?

:two: Could I leverage a storageclass hostpath in a PV that would allow me to make qovery databases run on a BYOK ? (not clear how to do so)

:three: Or is there another solution that eludes me ? :sweat_smile:

Thank you :pray: