Skip to content

Commit cd5cd4d

Browse files
Yu-Jackjillian-maroket
authored andcommitted
docs: revise based on feedback
Co-authored-by: Jillian Maroket <[email protected]> Signed-off-by: Jack Yu <[email protected]>
1 parent 612fa20 commit cd5cd4d

File tree

1 file changed

+7
-7
lines changed

1 file changed

+7
-7
lines changed

docs/advanced/addons/nvidiadrivertoolkit.md

Lines changed: 7 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -34,16 +34,16 @@ On pod startup, the ENTRYPOINT script will download the NVIDIA driver from the s
3434

3535
The `PCIDevices` addon can now leverage this addon to manage the lifecycle of the vGPU devices on nodes containing supported GPU [devices](../vgpusupport.md).
3636

37-
## Install Customized NVIDIA Driver
37+
## Install Different NVIDIA Driver Versions
3838

3939
_Available as of v1.7.0_
4040

41-
Harvester introduces a new feature that allows users to install different NVIDIA driver versions on different nodes.
41+
NVIDIA driver versions can vary across cluster nodes. If you want to install a specific driver version on a node, you must annotate the node before starting the nvidia-driver-toolkit add-on.
4242

43-
Before starting the nvidia-driver-toolkit, annotate the node where you want to install a specific driver version:
43+
```
44+
kubectl annotate nodes {node name} sriovgpu.harvesterhci.io/custom-driver=https://[driver location]
45+
```
4446

45-
- `kubectl annotate nodes {node name} sriovgpu.harvesterhci.io/custom-driver=https://[driver location]`
47+
The nvidia-driver-toolkit installs the specified driver version upon starting.
4648

47-
Then, start the nvidia-driver-toolkit. It will install the specified driver version.
48-
49-
For users who have previously installed an NVIDIA driver version, you need to restart the pod to trigger the installation process again.
49+
If an NVIDIA driver was previously installed, you must restart the pod to trigger the installation process again.

0 commit comments

Comments
 (0)