topic Re: csi driver error when migrating an OpenShift cluster from app-only to cloud-native full-stack mode in Container platforms

CSI driver error when migrating an OpenShift cluster from app-only to cloud-native full-stack mode

irivas1 — Thu, 30 Oct 2025 07:42:01 GMT

Hi,

We are trying to migrate an OpenShift cluster from application-only monitoring mode to cloud-native full-stack mode, but when doing so, we encountered this error:

FailedMount: MountVolume.SetUp failed for volume ‘osagent-storage’ : kubernetes.io/csi: mounter. SetUpAt failed to get CSI client: driver name csi.oneagent.dynatrace.com not found in the list of registered CSI drivers

It is as if the operator cannot install the CSI drivers in the worker nodes. If this is the case, it would be advisable to migrate to classic full-stack instead of cloud-native full-stack?

The requirement is monitor the Openshift routes and services that are hosted on the infra nodes using OneAgent ingestion.

Thanks in advance

Re: csi driver error when migrating an OpenShift cluster from app-only to cloud-native full-stack mode

Julius_Loman — Tue, 28 Oct 2025 18:50:41 GMT

Looks like your deployment misses the Dynatrace CSI driver. What exact steps did you do? How did you deploy the operator on particular? CSI is optional with applicationOnly but mandatory with cloudNativeFullstack.

Re: csi driver error when migrating an OpenShift cluster from app-only to cloud-native full-stack mode

irivas1 — Wed, 29 Oct 2025 07:17:52 GMT

Hi Julius,

Thanks for your answer.

When we deployed the operator, we did so in app-only mode and also deployed the CSI driver at the cluster level. The problem seems to be that when changing from monitoring mode to full-stack mode, since the csi driver is not installed on those nodes, this error occurs.

We also have some nodes isolated by taints, but we don't know where to apply the tolerations. In dynakube.yml, in csi-driver.yml?

Re: csi driver error when migrating an OpenShift cluster from app-only to cloud-native full-stack mode

g_kat — Wed, 29 Oct 2025 10:24:07 GMT

Hi,

Check this article: https://community.dynatrace.com/t5/Troubleshooting/Dynatrace-Operator-Having-a-different-number-of-OA-CSI-pods-than/ta-p/262171

It suggests to add the tolerations in the dynakube and csi files.

If you are using Helm, you can easily use the values.yaml(https://github.com/Dynatrace/dynatrace-operator/blob/v1.3.2/config/helm/chart/default/values.yaml) file to specify the necessary tolerations.

Re: csi driver error when migrating an OpenShift cluster from app-only to cloud-native full-stack mode

PacoPorro — Wed, 29 Oct 2025 13:36:57 GMT

Agree dynakube.yaml under cloud native full stack, add the tolerations.
In the openshift csi yaml , you should find it over here.

Re: csi driver error when migrating an OpenShift cluster from app-only to cloud-native full-stack mode

irivas1 — Wed, 29 Oct 2025 15:34:41 GMT

Hi,

Thanks for your answers.

If I understand correctly, in the dynakube.yml the tolerations should only go below line 10616. Is this correct?