feat(sandbox): switch device plugin to CDI injection mode#503
Open
feat(sandbox): switch device plugin to CDI injection mode#503
Conversation
elezar
force-pushed
the
feat/device-plugin-cdi-injection
branch
from
6840a21 to
6682d7d
Compare
Configure the NVIDIA device plugin to use deviceListStrategy=cdi-cri so that GPU devices are injected via direct CDI device requests in the CRI. Sandbox pods now only require the nvidia.com/gpu resource request — runtimeClassName is no longer set on GPU pods. Signed-off-by: Evan Lezar <elezar@nvidia.com>
Signed-off-by: Evan Lezar <elezar@nvidia.com>
elezar
force-pushed
the
feat/device-plugin-cdi-injection
branch
from
6682d7d to
9c39785
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Configure the NVIDIA device plugin to use
deviceListStrategy: cdi-criso GPU devices are injected via direct CDI device requests in the CRI. Sandbox pods now only neednvidia.com/gpu: 1in their resource limits —runtimeClassNameis no longer set on GPU pods.Related Issue
Related to #398
Changes
deploy/kube/gpu-manifests/nvidia-device-plugin-helmchart.yaml: adddeviceListStrategy: cdi-cri,cdi.nvidiaHookPath, andnvidiaDriverRoot: "/"to Helm valuescrates/openshell-server/src/sandbox/mod.rs: removeruntimeClassNameinsertion for GPU pods in bothsandbox_template_to_k8s()andinject_pod_template(); add unit test asserting CDI path sets noruntimeClassNamearchitecture/gateway-single-node.md: update GPU Enablement section to document CDI injection mode.agents/skills/debug-openshell-cluster/SKILL.md: add Step 8 with CDI-specific diagnostics (nvidia-ctk cdi list, device plugin logs, CDI spec files)Testing
mise run pre-commitpassesChecklist