GPU assigned to all pods of a chart, so the chart cannot be deployed. about charts

[Ornias1993]

We indeed do not support per-contain GPU selections on TrueNAS SCALE at this moment.
Thats not a bug, it's intended behavior.

Besides that I am going to close this, as this field is behavior like it should.
I do wonder what you need a GPU for on Nextcloud?
As far as I'm aware, we do not support anything Nextcloud related that can use GPU's.

[SBado]

I need to use harware acceleration for Memories, I also submitted a (merged) PR to add Intel drivers to the Nextcloud image: truecharts/containers#30296.

Also, everything was working fine until the last breaking changes in the GPU section, so I it seems to me that this use case was supported.

[Ornias1993]

I need to use harware acceleration for Memories, I also submitted a (merged) PR to add Intel drivers to the Nextcloud image: truecharts/containers#30296.

Also, everything was working fine until the last breaking changes in the GPU section, so I it seems to me that this use case was supported.

The GPU-section was a non-standard iX-Systems invention.
We're slowly removing any-and-all single-platform only code.

You're free to PR the required code to expose per-container resource sections, we're open to accepting such a PR.