[Bug 1249541] Issues while deploying the nvidia gpu operator following the official steps from: https://documentation.suse.com/cloudnative/rke2/latest/en/advanced.html#_operator_installation

[bugzilla_noreply at suse.com]

https://bugzilla.suse.com/show_bug.cgi?id=1249541
https://bugzilla.suse.com/show_bug.cgi?id=1249541#c6

Tomáš Bažant <tbazant at suse.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|CONFIRMED                   |IN_PROGRESS
                 CC|                            |tapas.nandi at suse.com
              Flags|                            |needinfo?(tapas.nandi at suse.
                   |                            |com)

--- Comment #6 from Tomáš Bažant <tbazant at suse.com> ---
comment from Siva Kanakala:

"The operator mounts the wrong folders into key pods ** because it calculated
them too early before reading thru all env settings.

transformForRuntime() is now called only after all env vars are finalized, so
volume mounts are generated from the final runtime configuration, not
partial/default values.

It seems Nvidia addressed this bug in GPU Operator v25.3.4. Please update your
GPU Operator to the latest version and see if that helps...

for more information read thru below: 

https://github.com/NVIDIA/gpu-operator/pull/1697/changes"

@Nandi, can we close this as it might not be doc issue in the end?
-- 
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.suse.com/pipermail/ai-maintainers/attachments/20260112/1ee25e71/attachment.htm>