add troubleshoot for GPU-Operator 25.10+ - #716
Conversation
Signed-off-by: Jifei Wang <jifei.wang@dynamia.ai>
✅ Deploy Preview for project-hami ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
📝 WalkthroughWalkthroughAdded English and Chinese troubleshooting documentation for HAMi Device Plugin and workload failures with NVIDIA GPU Operator 25.10+, including runtime, CDI, RuntimeClass, driver, injection, and Device Plugin configuration guidance. ChangesGPU Operator troubleshooting
Estimated code review effort: 2 (Simple) | ~15 minutes Possibly related PRs
Suggested labels: Suggested reviewers: 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
coderabbitai
Bot
added
the
kind/documentation
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/troubleshooting/troubleshooting.md`:
- Around line 122-126: Update the Helm value lookup in both troubleshooting
flows: docs/troubleshooting/troubleshooting.md lines 122-126 and
i18n/zh/docusaurus-plugin-content-docs/current/troubleshooting/troubleshooting.md
lines 114-118. Add the --all option to helm get values hami -n kube-system so
computed values, including default injection settings, are shown.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro Plus
Run ID: 9bb69af9-a19a-4f2d-a543-49579fc4987c
📒 Files selected for processing (4)
docs/faq/faq.mddocs/troubleshooting/troubleshooting.mdi18n/zh/docusaurus-plugin-content-docs/current/faq/faq.mdi18n/zh/docusaurus-plugin-content-docs/current/troubleshooting/troubleshooting.md
|
just 1 thing: |
spencercjh
left a comment
spencercjh
left a comment
There was a problem hiding this comment.
I think you need to copy these changes to the corresponding files in the v2.9.0 (latest stable version) directory.
Signed-off-by: Jifei Wang <jifei.wang@dynamia.ai>
There was a problem hiding this comment.
Actionable comments posted: 3
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@versioned_docs/version-v2.9.0/faq/faq.md`:
- Line 230: Update the FAQ entries at
versioned_docs/version-v2.9.0/faq/faq.md:230-230 and
i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/faq/faq.md:227-227 to
distinguish the hami-device-plugin monitor container port configured by
devicePlugin.monitorPort (9394 by default) from the Service/NodePort configured
by devicePlugin.service.httpPort (31992). State both ports clearly in English
and Chinese, using the Service port for Prometheus/scraping examples and the
monitor port as the Pod/container target.
In `@versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md`:
- Around line 61-72: Update the CDI and RuntimeClass verification commands in
the troubleshooting section of
versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md (lines 61-72)
and its Chinese counterpart
i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md
(lines 53-64) to validate spec.cdi.enabled=true and the nvidia RuntimeClass
.handler=nvidia using JSONPath, while preserving the existing HAMi Device Plugin
runtimeClass check.
- Around line 88-98: Correct the devicePlugin.nvidiaHookPath value from
nvidia-ctk to nvidia-cdi-hook in
versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md (lines 88-98)
and
i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md
(lines 80-90), preserving the existing GPU Operator path.
🪄 Autofix
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro Plus
Run ID: 84889e4e-1d55-4314-b963-b330534ed26b
📒 Files selected for processing (4)
i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/faq/faq.mdi18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.mdversioned_docs/version-v2.9.0/faq/faq.mdversioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md
There was a problem hiding this comment.
Actionable comments posted: 2
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (5)
versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md (3)
32-32: 🩺 Stability & Availability | 🟠 Major | ⚡ Quick winMake the Helm remediation explicit and reproducible.
The two one-line Helm upgrade commands silently fetch the latest chart and merge previous release values without showing the chart version. Pin the compatible chart version and add
--reuse-valuesin both remediation blocks.
versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md#L32-L32i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md#L27-L27🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md` at line 32, Update both Helm remediation commands in versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md:32-32 and i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md:27-27 to pin the compatible chart version and include --reuse-values, keeping the commands otherwise equivalent and synchronized.
18-18: 📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick winFix the broken prerequisites links in both copies.
The
installationdirectory is a sibling oftroubleshooting, so the current./installation/...path is incorrect.
versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md#L18-L18: Change the link to../installation/online-installation.i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md#L19-L19: Change the link to../installation/online-installation.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md` at line 18, Update the prerequisites link in versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md at lines 18-18 and i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md at lines 19-19 from ./installation/online-installation to ../installation/online-installation.
18-18: 🎯 Functional Correctness | 🟠 Major | ⚡ Quick winMake the
nvidiaruntime requirement conditional on non-CDI cases.GPU Operator 25.10+ enables CDI by default and standard Device Plugin workloads do not use
nvidiaas the container runtime by default. Keep thenvidiarequirement only for legacyenvvar/management-container paths, and update the matching Chinese text the same way.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md` at line 18, Update the runtime requirement text in versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md:18-18 to make nvidia required only for legacy envvar or management-container paths, not CDI or standard Device Plugin workloads. Apply the equivalent wording change in i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md:19-19.i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/faq/faq.md (2)
178-178: 🎯 Functional Correctness | 🟡 Minor | ⚡ Quick winFix the relative documentation links.
This file is under the
faq/directory. Therefore,./core-concepts/...,./userguide/..., and./developers/...resolve belowfaq/. Use../for sibling documentation directories.Proposed link corrections
- [GPU 虚拟化](./core-concepts/gpu-virtualization) + [GPU 虚拟化](../core-concepts/gpu-virtualization) - [动态 MIG 支持](./userguide/nvidia-device/dynamic-mig-support) + [动态 MIG 支持](../userguide/nvidia-device/dynamic-mig-support) - [GPU 利用率指标](./developers/gpu-utilization-metrics) + [GPU 利用率指标](../developers/gpu-utilization-metrics) - [Grafana Dashboard](./userguide/monitoring/grafana-dashboard) + [Grafana Dashboard](../userguide/monitoring/grafana-dashboard)Also applies to: 194-198, 221-221, 227-227
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/faq/faq.md` at line 178, Update the relative documentation links in the FAQ content, including the references near the cited sections, so sibling directories such as core-concepts, userguide, and developers use ../ instead of ./; preserve the linked targets and link text.
194-194: 📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick winImprove the Chinese wording for flexible memory sizing.
需要灵活的显存大小describes the memory size as “flexible”. Use a verb phrase that states the workload can configure memory flexibly.Proposed wording
-工作负载需要灵活的显存大小 +工作负载需要灵活地配置显存大小🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/faq/faq.md` at line 194, Update the Chinese wording in the HAMi vGPU usage sentence so “需要灵活的显存大小” is replaced with a verb phrase expressing that the workload can flexibly configure memory, while preserving the surrounding MIG comparison and link unchanged.Source: Linters/SAST tools
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@docs/faq/faq.md`:
- Line 181: Update the relative links in docs/faq/faq.md:181-181,
i18n/zh/docusaurus-plugin-content-docs/current/faq/faq.md:178-178, and
versioned_docs/version-v2.9.0/faq/faq.md:181-181 to use parent-relative paths
for core-concepts, userguide, and developers targets (../...). Apply the same
link corrections in all three FAQ copies.
In `@i18n/zh/docusaurus-plugin-content-docs/current/faq/faq.md`:
- Line 194: Update the sentence near the HAMi vGPU guidance to replace “灵活的显存大小”
with a grammatically correct modifier such as “灵活地调整显存大小,” while preserving the
surrounding meaning and structure.
---
Outside diff comments:
In `@i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/faq/faq.md`:
- Line 178: Update the relative documentation links in the FAQ content,
including the references near the cited sections, so sibling directories such as
core-concepts, userguide, and developers use ../ instead of ./; preserve the
linked targets and link text.
- Line 194: Update the Chinese wording in the HAMi vGPU usage sentence so
“需要灵活的显存大小” is replaced with a verb phrase expressing that the workload can
flexibly configure memory, while preserving the surrounding MIG comparison and
link unchanged.
In `@versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md`:
- Line 32: Update both Helm remediation commands in
versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md:32-32 and
i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md:27-27
to pin the compatible chart version and include --reuse-values, keeping the
commands otherwise equivalent and synchronized.
- Line 18: Update the prerequisites link in
versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md at lines 18-18
and
i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md
at lines 19-19 from ./installation/online-installation to
../installation/online-installation.
- Line 18: Update the runtime requirement text in
versioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md:18-18 to make
nvidia required only for legacy envvar or management-container paths, not CDI or
standard Device Plugin workloads. Apply the equivalent wording change in
i18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.md:19-19.
🪄 Autofix
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro Plus
Run ID: b02c1294-a744-4341-895e-b245435f77ae
📒 Files selected for processing (8)
docs/faq/faq.mddocs/troubleshooting/troubleshooting.mdi18n/zh/docusaurus-plugin-content-docs/current/faq/faq.mdi18n/zh/docusaurus-plugin-content-docs/current/troubleshooting/troubleshooting.mdi18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/faq/faq.mdi18n/zh/docusaurus-plugin-content-docs/version-v2.9.0/troubleshooting/troubleshooting.mdversioned_docs/version-v2.9.0/faq/faq.mdversioned_docs/version-v2.9.0/troubleshooting/troubleshooting.md
🚧 Files skipped from review as they are similar to previous changes (2)
- docs/troubleshooting/troubleshooting.md
- i18n/zh/docusaurus-plugin-content-docs/current/troubleshooting/troubleshooting.md
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: FouoF, rootsongjc, spencercjh The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
4 participants
What type of PR is this?
What this PR does / why we need it:
Which issue(s) this PR fixes:
Fixes #
Checklist:
npm run lintandnpm run format:checkpassnpm run buildsucceeds for bothenandzhgit commit -s)Summary by CodeRabbit
Documentation