1 Rookie

 • 

4 Posts

39

October 28th, 2025 16:16

PowerStore CSM documentation issue

From the documentation at https://dell.github.io/csm-docs/docs/getting-started/installation/openshift/powerstore/csmoperator/

 

In the setup of the CSM for the PowerStore there is a syntax error in the example that causes the deployment to fail with “Failed Prechecks: unsupported driver type”

This block is wrong, not only is the spacing out causing the failure but it is missing the previously defined secret.

cat << EOF > csm-powerstore.yaml

apiVersion: storage.dell.com/v1

kind: ContainerStorageModule

metadata:

  name: powerstore

  namespace: powerstore

spec:

driver:

   csiDriverType: "powerstore"

   configVersion: v2.15.0

   forceRemoveDriver: true

EOF

The correct example should be as follows.

cat << EOF > csm-powerstore.yaml

apiVersion: storage.dell.com/v1

kind: ContainerStorageModule

metadata:

  name: powerstore

  namespace: powerstore

spec:

  driver:

    csiDriverType: "powerstore"

    configVersion: v2.15.0

    forceRemoveDriver: true
    authSecret: powerstore-config

EOF

Please could you correct your documentation here (I'm not going to raise a support case for this and the dell.csm.operator@dell.com email address bounces). 


I will also say that IMO the example of the config secret should arguably use a blockProtocol type of “auto” rather than FC.

Moderator

 • 

14 Posts

October 29th, 2025 13:16

@Fastdruid​ - Thank you again for bringing the host NQN configuration issue to our attention. We’ve updated the documentation to restore the clarity that was previously available in version 1.13.2. The guidance around ensuring unique host NVMe Qualified Names (NQNs) for OpenShift or Kubernetes nodes connecting to Dell storage arrays has now been reintroduced into the upcoming version 1.15 refresh, scheduled for release this week. This includes the recommended MachineConfig to apply during deployment to ensure each node receives a unique NQN.

Once the update is live, you’ll be able to navigate to the relevant section using the left tree control via:
Getting Started → Installation → OpenShift → PowerStore → Operator → NVMeFC or NVMeTCP tabs.

We truly appreciate your diligence in identifying this gap and helping us improve the deployment experience for all users.

Moderator

 • 

14 Posts

October 28th, 2025 18:19

@Fastdruid  - Thank you for taking the time to share your feedback with us. We truly appreciate your engagement and the insights you've provided. We recognize the issues you've highlighted in the documentation and want to assure you that we are actively working to address them. Your input helps us improve, and we're grateful for your contribution to the community.

Could you please clarify the origin or intended use of the email address dell.csm.operator@dell.com

Again, thank you for your help and input!

1 Rookie

 • 

4 Posts

October 28th, 2025 18:41

@aron_dell​ 

Finding _where_ to report issues with the docs is really hard. The github has issues disabled. There are no feedback options on the docs themselves. 

That email is given in the Red Hat docs as the contact for support for the CSM 

https://catalog.redhat.com/en/software/container-stacks/detail/639c131cd113cc28c0c2f2d5


Moderator

 • 

2 Posts

October 28th, 2025 18:56

@Fastdruid , are you getting any errors if you exclude the "authSecret: powerstore-config" parameter from the CSM CR?

Moderator

 • 

14 Posts

October 28th, 2025 19:17

@Fastdruid​ - We’ve worked with our RedHat partner to resolve the support linking issue. Customers are now directed to our support website, where they can choose to open a ticket or engage with the Community as an alternative. We appreciate your patience and continued support!

1 Rookie

 • 

4 Posts

October 28th, 2025 19:35

@TrevorDawe​ Didn't try.  it shouldn't as it should auto-generate from <metadataName>-config, if someone however has used a different name for the secret it will be obvious at that point it needs changing. 

 




(edited)

1 Rookie

 • 

4 Posts

October 28th, 2025 19:38

While we're on issues with the docs. 


Version 1.13.2 of the docs mentions

  • All OpenShift or Kubernetes nodes connecting to Dell storage arrays must use unique host NVMe Qualified Names (NQNs).

The OpenShift deployment process for CoreOS will set the same host NQN for all nodes. The host NQN is stored in the file /etc/nvme/hostnqn. One possible solution to ensure unique host NQNs is to add the following machine config to your OCP cluster:



https://dell.github.io/csm-docs/v2/deployment/helm/drivers/installation/powerstore/

This is _NOT_ mentioned in 1.15 (or the MachineConfig to reset it given), when installing for OpenShift, so everything gets the same NQN and kind of works, kind of fails. 

(edited)

Moderator

 • 

2 Posts

October 29th, 2025 13:18

@Fastdruid​ Ok great, I just wanted to confirm in case you were seeing any issues with authSecret. Yes, it will only be required if using a different name.

(edited)

Moderator

 • 

14 Posts

October 29th, 2025 14:25

@Fastdruid - If everything looks good and your query has been fully addressed, please feel free to mark it as solved by selecting "Mark as Accepted Answer." This helps us close the loop and ensures others can benefit from the resolution as well.

Moderator

 • 

14 Posts

October 30th, 2025 15:20

@Fastdruid - Thank you again for your engagement and collaboration throughout this query. As we haven’t received a follow-up and the issue appears to be fully addressed, we’ll go ahead and mark this as solved to help close the loop and ensure others can benefit from the resolution. Your feedback and responsiveness are greatly appreciated — they help us continuously improve the experience for all our customers.

No Events found!

Top