NVIDIA-918: Add NVIDIA DPF Operator documentation - #117142
Conversation
|
@gaurav-nelson: This pull request references NVIDIA-918 which is a valid jira issue. Warning: The referenced jira issue has an invalid target version for the target branch this PR targets: expected the epic to target the "5.0.0" version, but no target version was set. DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
e112c56 to
3e87522
Compare
|
👋 @gaurav-nelson Great work on the comprehensive DPF documentation! I noticed some production-critical gaps that would prevent successful DTS observability deployment, so I created PR #117216 with targeted additions: 🚨 Critical Missing Piece: Complete Grafana installation instructions
📋 Other Additions:
All content from production-tested deployment, YAML validated with Happy to collaborate on merging these fixes into your PR instead of having competing PRs. The additions complement your excellent work perfectly! View specific additions: main...szigmon:openshift-docs:dpf-dts-critical-fixes-20260804 What do you think? 🤝 |
🚀 DPF Documentation Enhancement - How Would You Like to Proceed?Hi @gaurav-nelson! Excellent comprehensive DPF documentation work! I've identified critical production gaps that would prevent successful DTS observability deployment and prepared comprehensive fixes. How would you prefer to receive these enhancements? 🎯 Critical Issues Found
🤝 Integration Options - Your ChoiceOption A: Direct File Sharing ⭐ (Easiest)I provide ready-to-use files via:
Option B: Collaborative PR 🔄 (Most Professional)
Option C: Detailed Instructions 📋 (Complete Control)
Option D: Live Collaboration 💬 (Interactive)
📊 What's ReadyComplete content prepared (production-tested):
Total: 781 lines of comprehensive, production-ready content 🎯 My RecommendationOption B (Collaborative PR) because:
🤔 What's Your Preference?Just let me know which option works best for you! I'm happy to accommodate whatever approach is most convenient for your workflow. Goal: Ensure users can successfully deploy DTS observability in production environments 🚀 All content is based on working production deployment and addresses real deployment blockers I've encountered. |
JoeAldinger
left a comment
There was a problem hiding this comment.
A few comments and suggestions. Overall there are some global changes I'd recommend to improve the user experience, increase readability, and condense the docs. Thanks, and let me know if you have any questions.
| .Additional resources | ||
|
|
||
| * xref:../dpu-operator/dpu-operator.adoc#dpu-operator[DPU Operator] | ||
| * link:https://github.com/NVIDIA/doca-platform/tree/public-release-v26.4[DOCA Platform Framework (DPF) GitHub] |
There was a problem hiding this comment.
You'll need approval to link to GH repos. cc @kalexand-rh
|
|
||
| In Grafana, go to *Dashboards* and open *DOCA DPU Telemetry (DTS)*. | ||
| Use the *DPU (source)* dropdown to focus on a specific DPU or select *All*. | ||
| Adjust the time range by using the time-range control on the dashboard toolbar. The dashboard refreshes every 30 seconds. |
There was a problem hiding this comment.
| Adjust the time range by using the time-range control on the dashboard toolbar. The dashboard refreshes every 30 seconds. | |
| Adjust the time range by using the *time-range* control on the dashboard toolbar. The dashboard refreshes every 30 seconds. |
There was a problem hiding this comment.
I'm guessing this is a GUI element and is hyphenated like that.
| . Label each worker node for the DPU `MachineConfigPool`: | ||
| + | ||
| In the following command, replace `<worker_node_name>` with the name of the worker node: | ||
| + |
There was a problem hiding this comment.
Replace <worker_node_name> with the name of the worker node.
+
JoeAldinger
left a comment
There was a problem hiding this comment.
A few comments and suggestions. Overall there are some global changes I'd recommend to improve the user experience, increase readability, and condense the docs. Thanks, and let me know if you have any questions.
|
@szigmon I cant find the release notes here, are they around? |
working on a new PR with all the missing additions |
3e87522 to
fc28949
Compare
…penshift#117142 This commit adds comprehensive enhancements to the NVIDIA DPF Operator documentation, ensuring full alignment with PR openshift#117142 while providing significant production-ready improvements for OpenShift v4.22 deployments. MAJOR ENHANCEMENTS: • Native OpenShift Console DTS integration (no Grafana dependency) • Comprehensive troubleshooting framework (8 detailed modules) • Enhanced release notes for v26.4.1-beta.1 • Complete traffic validation testing procedures NEW MODULES ADDED (9 files): • nw-dpf-installing-dts-console-dashboard.adoc - Native console telemetry • nw-dpf-troubleshooting-framework.adoc - Main troubleshooting guide • nw-dpf-troubleshooting-dpu-provisioning.adoc - DPU provisioning issues • nw-dpf-troubleshooting-hosted-cluster.adoc - HyperShift cluster issues • nw-dpf-troubleshooting-networking.adoc - OVN-K and networking issues • nw-dpf-troubleshooting-diagnostics.adoc - Diagnostic commands reference • nw-dpf-ts-bmc-certificates.adoc - BMC certificate troubleshooting • nw-dpf-ts-dpu-node-status.adoc - DPU node NotReady resolution • nw-dpf-ts-worker-csr-approval.adoc - CSR approval failure diagnostics CRITICAL FIXES: • Fixed HOSTED_CLUSTER_VIP: 203.0.113.200 → 192.168.1.200 • Added missing traffic validation tests (NodePort connectivity) • Enhanced release notes with current version information • Updated assemblies for proper module integration ALIGNMENT WITH PR openshift#117142: • Verified compatibility with all 69 files from tech writers' PR • Added missing troubleshooting modules referenced in original PR • Enhanced beyond original scope for production readiness • Maintained Red Hat documentation standards throughout TECHNICAL IMPROVEMENTS: • All environment variables use proper $VARIABLE syntax • Spoofed hostnames replace real server names for security • Complete diagnostic command reference with expected outputs • Professional troubleshooting workflows for enterprise deployments Total: 5 modified + 9 new files = 14 files enhanced Coverage: Complete Chapters 1-11 documentation with production-ready troubleshooting Co-authored-by: Claude Sonnet 4 <noreply@anthropic.com>
|
@gaurav-nelson Updated enhancement, aligned with your latest revision (fc28949): Single commit, 14 files — view diff New modules: console DTS dashboard, comprehensive troubleshooting framework (5 modules), diagnostic commands. Branch: |
PR was updated Aug 7 after tech writer review (JoeAldinger): - Removed 18 modules (consolidated operators, removed overview wrappers) - Added grafana-for-dts, release-notes, ts-bmc/dpu/csr modules This commit fixes 3 alignment issues: - dpf-validation-telemetry: match flat +1 structure, remove deleted module refs (validating-traffic-overview, configuring-dpu-telemetry), remove release-notes (stays in about-dpf-operator per PR), add :dpf-version: attribute - dpf-troubleshooting: restore PR's original ts- modules first, then append our new comprehensive modules, remove unsupported-ovnk ref, add :dpf-version: attribute - release-notes: fix assembly comment to about-dpf-operator.adoc Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Aligned with PR openshift#117142 latest revision (fc28949, Aug 7). New modules (6): - Console DTS dashboard with native OpenShift integration - Comprehensive troubleshooting framework - DPU provisioning troubleshooting - Hosted cluster troubleshooting - Networking troubleshooting - Diagnostic commands and log collection Enhanced modules (6): - release-notes: full v26.4.1-beta.1 release notes, unsupported OVN-K features, expanded known issues (179 lines, up from 105) - running-traffic-validation: added worker-to-worker service and optional NodePort tests - hcp-environment-variables: fixed VIP to 192.168.1.200, added export commands block - ts-bmc-certificates: expanded BMC troubleshooting (53 lines) - ts-dpu-node-status: expanded DPU node diagnostics (84 lines) - ts-worker-csr-approval: expanded CSR troubleshooting (86 lines) Assembly updates (2): - dpf-validation-telemetry: added console dashboard include - dpf-troubleshooting: added 5 comprehensive troubleshooting modules after existing ts- modules Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
fc28949 to
3a1e992
Compare
|
/retest |
cad07fb to
273418a
Compare
JoeAldinger
left a comment
There was a problem hiding this comment.
A few suggestions mostly the new content has a good bit of self referential language in the abstracts that needs to be removed. I left a question about the release notes module that might be worth discussing with engineering.
| Connection to 172.30.102.123 5000 port [tcp/*] succeeded! | ||
| ---- | ||
| + | ||
| A `succeeded` message confirms that the service is reachable through the DPU-accelerated network. |
There was a problem hiding this comment.
IMO, you don't need the actual output if you are going to describe what the user should be looking for, for example A succeeded message...
273418a to
0e622ef
Compare
7e6d10b to
13dffef
Compare
Co-authored-by: Cursor <cursoragent@cursor.com>
13dffef to
65248e6
Compare
|
@gaurav-nelson: all tests passed! Full PR test history. Your PR dashboard. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here. |
Based on #117002 by @Shivank-rh
Cherrypick:
Summary
networking/networking_operators/dpf_operator/, 59 modules inmodules/, 2 images, topic map entryStructure
about-dpf-operatordpf-environment-setupdpf-operator-installationdpf-hosted-cluster-provisioningdpf-worker-nodes-provisioningdpf-validation-telemetrydpf-troubleshootingItems for reviewer
xref:placeholders — need anchor ID verificationTest plan
include::paths resolve🤖 Generated with Claude Code